From 607e4a4e30925c792bc2a23f58766a2cd7652ed9 Mon Sep 17 00:00:00 2001 From: Tianhe Zhang Date: Tue, 18 Aug 2026 14:16:08 -0700 Subject: [PATCH 1/9] feat(spend-logs): add lifecycle timestamps --- .../migration.sql | 3 +++ .../litellm_proxy_extras/schema.prisma | 2 ++ litellm/models/spend_logs.py | 2 ++ litellm/proxy/schema.prisma | 2 ++ tests/test_litellm/models/test_models.py | 19 +++++++++++++++++++ 5 files changed, 28 insertions(+) create mode 100644 litellm-proxy-extras/litellm_proxy_extras/migrations/20260818000000_add_spend_log_timestamps/migration.sql diff --git a/litellm-proxy-extras/litellm_proxy_extras/migrations/20260818000000_add_spend_log_timestamps/migration.sql b/litellm-proxy-extras/litellm_proxy_extras/migrations/20260818000000_add_spend_log_timestamps/migration.sql new file mode 100644 index 00000000000..a4a3cc3bb1b --- /dev/null +++ b/litellm-proxy-extras/litellm_proxy_extras/migrations/20260818000000_add_spend_log_timestamps/migration.sql @@ -0,0 +1,3 @@ +ALTER TABLE "LiteLLM_SpendLogs" +ADD COLUMN IF NOT EXISTS "created_at" TIMESTAMP(3) NOT NULL DEFAULT CURRENT_TIMESTAMP, +ADD COLUMN IF NOT EXISTS "updated_at" TIMESTAMP(3) NOT NULL DEFAULT CURRENT_TIMESTAMP; diff --git a/litellm-proxy-extras/litellm_proxy_extras/schema.prisma b/litellm-proxy-extras/litellm_proxy_extras/schema.prisma index 24c0f1f11cc..52fb447157b 100644 --- a/litellm-proxy-extras/litellm_proxy_extras/schema.prisma +++ b/litellm-proxy-extras/litellm_proxy_extras/schema.prisma @@ -641,6 +641,8 @@ model LiteLLM_SpendLogs { mcp_namespaced_tool_name String? agent_id String? proxy_server_request Json? @default("{}") + created_at DateTime @default(now()) @map("created_at") + updated_at DateTime @default(now()) @updatedAt @map("updated_at") @@index([startTime]) @@index([startTime, request_id]) @@index([end_user]) diff --git a/litellm/models/spend_logs.py b/litellm/models/spend_logs.py index c5a0522864a..92b1a753ad5 100644 --- a/litellm/models/spend_logs.py +++ b/litellm/models/spend_logs.py @@ -33,6 +33,8 @@ class LiteLLM_SpendLogs(LiteLLMPydanticObjectBase): requester_ip_address: str | None = None messages: str | list | dict | None response: str | list | dict | None + created_at: datetime | None = None + updated_at: datetime | None = None class LiteLLM_ErrorLogs(LiteLLMPydanticObjectBase): diff --git a/litellm/proxy/schema.prisma b/litellm/proxy/schema.prisma index 24c0f1f11cc..52fb447157b 100644 --- a/litellm/proxy/schema.prisma +++ b/litellm/proxy/schema.prisma @@ -641,6 +641,8 @@ model LiteLLM_SpendLogs { mcp_namespaced_tool_name String? agent_id String? proxy_server_request Json? @default("{}") + created_at DateTime @default(now()) @map("created_at") + updated_at DateTime @default(now()) @updatedAt @map("updated_at") @@index([startTime]) @@index([startTime, request_id]) @@index([end_user]) diff --git a/tests/test_litellm/models/test_models.py b/tests/test_litellm/models/test_models.py index 786f6244930..187c7aa7f5e 100644 --- a/tests/test_litellm/models/test_models.py +++ b/tests/test_litellm/models/test_models.py @@ -498,6 +498,25 @@ class TestSpendLogs: assert log.request_id == "r1" assert log.spend == 0.0 assert log.cache_hit == "False" + assert log.created_at is None + assert log.updated_at is None + + def test_spend_logs_parse_database_timestamps(self): + created_at = datetime(2026, 8, 18, 12, 0, 0) + updated_at = datetime(2026, 8, 18, 12, 5, 0) + log = LiteLLM_SpendLogs( + request_id="r1", + api_key="sk-1", + call_type="completion", + startTime=None, + endTime=None, + messages=None, + response=None, + created_at=created_at, + updated_at=updated_at, + ) + assert log.created_at == created_at + assert log.updated_at == updated_at def test_error_logs_creation(self): log = LiteLLM_ErrorLogs( From 4f4892efcbf56d324ad8339f29a87f7dd6dd663a Mon Sep 17 00:00:00 2001 From: Tianhe Zhang Date: Tue, 18 Aug 2026 14:29:02 -0700 Subject: [PATCH 2/9] fix(spend-logs): sync root prisma schema --- schema.prisma | 2 ++ 1 file changed, 2 insertions(+) diff --git a/schema.prisma b/schema.prisma index 24c0f1f11cc..52fb447157b 100644 --- a/schema.prisma +++ b/schema.prisma @@ -641,6 +641,8 @@ model LiteLLM_SpendLogs { mcp_namespaced_tool_name String? agent_id String? proxy_server_request Json? @default("{}") + created_at DateTime @default(now()) @map("created_at") + updated_at DateTime @default(now()) @updatedAt @map("updated_at") @@index([startTime]) @@index([startTime, request_id]) @@index([end_user]) From 6acf970009ba53a18203f23b8d3711a352a90718 Mon Sep 17 00:00:00 2001 From: mateo-berri <277851410+mateo-berri@users.noreply.github.com> Date: Wed, 19 Aug 2026 12:41:52 -0700 Subject: [PATCH 3/9] chore(ui): regenerate dashboard API types for spend log timestamps --- ui/litellm-dashboard/src/lib/http/schema.d.ts | 4 ++++ 1 file changed, 4 insertions(+) diff --git a/ui/litellm-dashboard/src/lib/http/schema.d.ts b/ui/litellm-dashboard/src/lib/http/schema.d.ts index 3a129890267..f2c6b978e39 100644 --- a/ui/litellm-dashboard/src/lib/http/schema.d.ts +++ b/ui/litellm-dashboard/src/lib/http/schema.d.ts @@ -27684,6 +27684,8 @@ export interface components { * @default 0 */ completion_tokens: number | null; + /** Created At */ + created_at?: string | null; /** Endtime */ endTime: string | null; /** Messages */ @@ -27727,6 +27729,8 @@ export interface components { * @default 0 */ total_tokens: number | null; + /** Updated At */ + updated_at?: string | null; /** * User * @default From 73e7105e606edf8b6770892f2ef026417a76ab79 Mon Sep 17 00:00:00 2001 From: ryan-crabbe-berri Date: Wed, 19 Aug 2026 13:28:20 -0700 Subject: [PATCH 4/9] fix(ui): restore tab strip styling and panel persistence lost in the shadcn migration (#37403) Tremor's TabList defaulted to the underlined `line` variant and its TabPanel rendered every panel, hiding the inactive ones with a class. The shadcn TabsList defaults to a segmented pill and Base UI's TabsPanel unmounts a hidden panel unless it carries `keepMounted`, so the conversion waves quietly changed both on the pages that took tremor's defaults. Restores the underline on the nine strips whose tremor markup carried no `variant`, leaving the ones that were explicitly `variant="solid"` as pills, and puts `keepMounted` back on the thirteen files whose panels used to stay mounted, so filters, scroll position and in-progress input survive a tab switch again. Seeding the old usage page's activity state properly comes with it: it was cast from `{}`, so the panel crashed on `data.length` the moment it mounted before its fetch resolved, which only stayed hidden while the panel was unmounted until first opened. --- .../_components/APIReferenceView.tsx | 6 ++-- .../budgets/_components/budget_panel.tsx | 10 +++--- .../caching/_components/cache_dashboard.tsx | 20 +++++------ .../caching/_components/cache_health.tsx | 4 +-- .../_components/cost_tracking_settings.tsx | 14 +++++--- .../_components/mcp_server_view.tsx | 14 ++++---- .../mcp-servers/_components/mcp_servers.tsx | 12 +++---- .../old-usage/_components/usage.tsx | 16 +++++---- .../src/app/(dashboard)/playground/page.tsx | 12 ++++--- .../policies/_components/index.tsx | 18 +++++----- .../_components/general_settings.tsx | 10 +++--- .../components/AIHub/ModelHubTable.test.tsx | 36 +++++++++++++++++++ .../src/components/AIHub/ModelHubTable.tsx | 26 +++++++++----- .../src/components/per_user_usage.tsx | 6 ++-- .../src/components/settings.tsx | 10 +++--- .../components/templates/key_info_view.tsx | 10 ++++-- .../src/components/user_agent_activity.tsx | 14 ++++---- 17 files changed, 148 insertions(+), 90 deletions(-) diff --git a/ui/litellm-dashboard/src/app/(dashboard)/api-reference/_components/APIReferenceView.tsx b/ui/litellm-dashboard/src/app/(dashboard)/api-reference/_components/APIReferenceView.tsx index 9342017ed3f..847bd34da3e 100644 --- a/ui/litellm-dashboard/src/app/(dashboard)/api-reference/_components/APIReferenceView.tsx +++ b/ui/litellm-dashboard/src/app/(dashboard)/api-reference/_components/APIReferenceView.tsx @@ -46,7 +46,7 @@ const APIReferenceView: React.FC = ({ proxySettings }) => { Langchain Py - + - + - + = ({ accessToken }) => { - +
{selectedBudget && ( @@ -134,7 +134,7 @@ const BudgetPanel: React.FC = ({ accessToken }) => { />
- +

How to use budget id

@@ -149,13 +149,13 @@ const BudgetPanel: React.FC = ({ accessToken }) => { Test it (OpenAI SDK) - + {CREATE_END_USER_CURL_COMMAND} - + {CHAT_COMPLETIONS_CURL_COMMAND} - + {OPENAI_SDK_PYTHON_CODE} diff --git a/ui/litellm-dashboard/src/app/(dashboard)/caching/_components/cache_dashboard.tsx b/ui/litellm-dashboard/src/app/(dashboard)/caching/_components/cache_dashboard.tsx index 62c456b1b38..cafa7a2ce37 100644 --- a/ui/litellm-dashboard/src/app/(dashboard)/caching/_components/cache_dashboard.tsx +++ b/ui/litellm-dashboard/src/app/(dashboard)/caching/_components/cache_dashboard.tsx @@ -140,18 +140,18 @@ const CacheDashboard: React.FC = ({ accessToken, token, userRole return ( -
- - +
+ + Cache Analytics - + Cache Health - + Cache Settings - + Coordination Redis @@ -164,7 +164,7 @@ const CacheDashboard: React.FC = ({ accessToken, token, userRole
- +

@@ -311,7 +311,7 @@ const CacheDashboard: React.FC = ({ accessToken, token, userRole - + = ({ accessToken, token, userRole /> - + - + diff --git a/ui/litellm-dashboard/src/app/(dashboard)/caching/_components/cache_health.tsx b/ui/litellm-dashboard/src/app/(dashboard)/caching/_components/cache_health.tsx index 0f52bed874d..221a17f911d 100644 --- a/ui/litellm-dashboard/src/app/(dashboard)/caching/_components/cache_health.tsx +++ b/ui/litellm-dashboard/src/app/(dashboard)/caching/_components/cache_health.tsx @@ -164,7 +164,7 @@ const HealthCheckDetails: React.FC<{ response: any }> = ({ response }) => { - +

{response?.status === "healthy" ? ( @@ -228,7 +228,7 @@ const HealthCheckDetails: React.FC<{ response: any }> = ({ response }) => {
- + 
               {(() => {
diff --git a/ui/litellm-dashboard/src/app/(dashboard)/cost-tracking/_components/cost_tracking_settings.tsx b/ui/litellm-dashboard/src/app/(dashboard)/cost-tracking/_components/cost_tracking_settings.tsx
index 029f8aa424c..85350a539c2 100644
--- a/ui/litellm-dashboard/src/app/(dashboard)/cost-tracking/_components/cost_tracking_settings.tsx
+++ b/ui/litellm-dashboard/src/app/(dashboard)/cost-tracking/_components/cost_tracking_settings.tsx
@@ -197,11 +197,15 @@ const CostTrackingSettings: React.FC = ({ userID, use
             />
             
               
-                
-                  Discounts
-                  Test It
+                
+                  
+                    Discounts
+                  
+                  
+                    Test It
+                  
                 
-                
+                
                   

                     

                       
@@ -237,7 +241,7 @@ const CostTrackingSettings: React.FC = ({ userID, use
                     )}
                   

                 
-                
+                
                   

                     
                   

diff --git a/ui/litellm-dashboard/src/app/(dashboard)/mcp-servers/_components/mcp_server_view.tsx b/ui/litellm-dashboard/src/app/(dashboard)/mcp-servers/_components/mcp_server_view.tsx
index 736327409d6..9c5396d1e9a 100644
--- a/ui/litellm-dashboard/src/app/(dashboard)/mcp-servers/_components/mcp_server_view.tsx
+++ b/ui/litellm-dashboard/src/app/(dashboard)/mcp-servers/_components/mcp_server_view.tsx
@@ -130,22 +130,22 @@ export const MCPServerView: React.FC = ({
       

 
        setSelectedTabIndex(Number(v))}>
-        
-          
+        
+          
             Overview
           
-          
+          
             MCP Tools
           
           {isProxyAdmin && (
-            
+            
               Settings
             
           )}
         
 
         {/* Overview Panel */}
-        
+        
           

             
               
Transport

@@ -192,7 +192,7 @@ export const MCPServerView: React.FC = ({
         
 
         {/* Tool Panel */}
-        
+        
            = ({
         
 
         {/* Settings Panel */}
-        
+        
           
             

               
MCP Server Settings

diff --git a/ui/litellm-dashboard/src/app/(dashboard)/mcp-servers/_components/mcp_servers.tsx b/ui/litellm-dashboard/src/app/(dashboard)/mcp-servers/_components/mcp_servers.tsx
index 970fb56799d..ed0c77064b9 100644
--- a/ui/litellm-dashboard/src/app/(dashboard)/mcp-servers/_components/mcp_servers.tsx
+++ b/ui/litellm-dashboard/src/app/(dashboard)/mcp-servers/_components/mcp_servers.tsx
@@ -543,7 +543,7 @@ const MCPServers: React.FC = ({ accessToken, userRole, userID })
               
             )}
           
-          
+          
             {selectedServerId ? (
                = ({ accessToken, userRole, userID })
               

             )}
           
-          
+          
             
           
-          
+          
             
           
           {isAdminRole(userRole) && (
-            
+            
               
             
           )}
           {isAdminRole(userRole) && (
-            
+            
               
             
           )}
           {isAdminRole(userRole) && (
-            
+            
               
             
           )}
diff --git a/ui/litellm-dashboard/src/app/(dashboard)/old-usage/_components/usage.tsx b/ui/litellm-dashboard/src/app/(dashboard)/old-usage/_components/usage.tsx
index 7fe67fb9a41..aca1db0fbae 100644
--- a/ui/litellm-dashboard/src/app/(dashboard)/old-usage/_components/usage.tsx
+++ b/ui/litellm-dashboard/src/app/(dashboard)/old-usage/_components/usage.tsx
@@ -56,6 +56,8 @@ interface GlobalActivityData {
   daily_data: { date: string; api_requests: number; total_tokens: number }[];
 }
 
+const EMPTY_GLOBAL_ACTIVITY: GlobalActivityData = { sum_api_requests: 0, sum_total_tokens: 0, daily_data: [] };
+
 type UsageDateRange = { from?: Date; to?: Date };
 
 type TeamSpendTotal = { name: string; value: number };
@@ -105,7 +107,7 @@ const UsagePage: React.FC = ({ accessToken, token, userRole, use
   const [uniqueTeamIds, setUniqueTeamIds] = useState([]);
   const [totalSpendPerTeam, setTotalSpendPerTeam] = useState([]);
   const [spendByProvider, setSpendByProvider] = useState([]);
-  const [globalActivity, setGlobalActivity] = useState({} as GlobalActivityData);
+  const [globalActivity, setGlobalActivity] = useState(EMPTY_GLOBAL_ACTIVITY);
   const [globalActivityPerModel, setGlobalActivityPerModel] = useState([]);
   const [selectedKeyToken, setSelectedKeyToken] = useState(null);
   const [selectedTags, setSelectedTags] = useState([ALL_TAGS]);
@@ -560,14 +562,14 @@ const UsagePage: React.FC = ({ accessToken, token, userRole, use
           )}
         
 
-        
+        
           
             
               Cost
               Activity
             
 
-            
+            
               

                 

                   

@@ -671,7 +673,7 @@ const UsagePage: React.FC = ({ accessToken, token, userRole, use
               

             
 
-            
+            
               

                 
                   
@@ -751,7 +753,7 @@ const UsagePage: React.FC = ({ accessToken, token, userRole, use
           
         
 
-        
+        
           

             

               
@@ -782,7 +784,7 @@ const UsagePage: React.FC = ({ accessToken, token, userRole, use
           

         
 
-        
+        
           

             Customers of your LLM API calls. Tracked when a `user` param is passed in your LLM calls{" "}
              = ({ accessToken, token, userRole, use
           
         
 
-        
+        
           

             

               
         
-        
+        
           
         
-        
+        
           
         
-        
+        
           
         
-        
+        
           
            = ({ accessToken, userRole })
   return (
     

       
-        
-          
+        
+          
             Templates
           
-          
+          
             Policies
           
-          
+          
             Attachments
           
-          
+          
             Policy Simulator
           
         
 
-        
+        
           
            = ({ accessToken, userRole })
           />
         
 
-        
+        
           
 
           

@@ -503,7 +503,7 @@ const PoliciesPanel: React.FC = ({ accessToken, userRole })
           />
         
 
-        
+        
           }>
             

               Policy attachments control where your policies apply. Policies don't do anything until you attach
@@ -571,7 +571,7 @@ const PoliciesPanel: React.FC = ({ accessToken, userRole })
           />
         
 
-        
+        
           
         
       
diff --git a/ui/litellm-dashboard/src/app/(dashboard)/router-settings/_components/general_settings.tsx b/ui/litellm-dashboard/src/app/(dashboard)/router-settings/_components/general_settings.tsx
index 2a5b94b2fd7..1e58e81f4e3 100644
--- a/ui/litellm-dashboard/src/app/(dashboard)/router-settings/_components/general_settings.tsx
+++ b/ui/litellm-dashboard/src/app/(dashboard)/router-settings/_components/general_settings.tsx
@@ -261,19 +261,19 @@ const GeneralSettings: React.FC = ({ accessToken, user
           Prompt Caching
           General
         
-        
+        
           
         
-        
+        
           
         
-        
+        
           
         
-        
+        
           
         
-        
+        
           
             
               
diff --git a/ui/litellm-dashboard/src/components/AIHub/ModelHubTable.test.tsx b/ui/litellm-dashboard/src/components/AIHub/ModelHubTable.test.tsx
index 5c4fa4eed43..397e2ab10fa 100644
--- a/ui/litellm-dashboard/src/components/AIHub/ModelHubTable.test.tsx
+++ b/ui/litellm-dashboard/src/components/AIHub/ModelHubTable.test.tsx
@@ -1,4 +1,5 @@
 import * as networking from "@/components/networking";
+import userEvent from "@testing-library/user-event";
 import { afterEach, beforeEach, describe, expect, it, vi } from "vitest";
 import { renderWithProviders, screen, waitFor } from "../../../tests/test-utils";
 import ModelHubTable from "./ModelHubTable";
@@ -201,6 +202,41 @@ describe("ModelHubTable", () => {
     expect(getUiConfigCallOrder).toBeLessThan(modelHubPublicModelsCallOrder);
   });
 
+  describe("hub tabs", () => {
+    const renderHub = async () => {
+      vi.mocked(networking.modelHubCall).mockResolvedValue({
+        data: [{ model_group: "claude-opus-4-8", providers: ["anthropic"], mode: "chat" }],
+      });
+      vi.mocked(networking.getConfigFieldSetting).mockResolvedValue({ field_value: false });
+      vi.mocked(networking.getAgentsList).mockResolvedValue({ agents: [] });
+      vi.mocked(networking.fetchMCPServers).mockResolvedValue([]);
+      vi.mocked(networking.getUiSettings).mockResolvedValue({ values: {} });
+      mockUseUISettings.mockReturnValue({ data: { values: {} }, isLoading: false });
+
+      const user = userEvent.setup();
+      renderWithProviders(
+        ,
+      );
+      return { user, search: await screen.findByPlaceholderText("Search model names...") };
+    };
+
+    it("keeps the model filter typed on the Model Hub tab after visiting another hub", async () => {
+      const { user, search } = await renderHub();
+
+      await user.type(search, "opus");
+      await user.click(screen.getByRole("tab", { name: "Agent Hub" }));
+      await user.click(screen.getByRole("tab", { name: "Model Hub" }));
+
+      expect(await screen.findByPlaceholderText("Search model names...")).toHaveValue("opus");
+    });
+
+    it("renders the hub strip as underlined tabs rather than a segmented pill", async () => {
+      await renderHub();
+
+      expect(screen.getByRole("tablist")).toHaveAttribute("data-variant", "line");
+    });
+  });
+
   describe("authentication redirect behavior", () => {
     // Test cases where requireAuth is true - should redirect on invalid tokens
     testAuthRedirect(
diff --git a/ui/litellm-dashboard/src/components/AIHub/ModelHubTable.tsx b/ui/litellm-dashboard/src/components/AIHub/ModelHubTable.tsx
index 28dac9badcd..abddb9577e0 100644
--- a/ui/litellm-dashboard/src/components/AIHub/ModelHubTable.tsx
+++ b/ui/litellm-dashboard/src/components/AIHub/ModelHubTable.tsx
@@ -428,16 +428,24 @@ const ModelHubTable: React.FC = ({ accessToken, publicPage,
 
           {/* Tab System for Model Hub, Agent Hub, MCP Hub, and Plugin Marketplace */}
           
-            
-              Model Hub
-              Agent Hub
-              MCP Hub
-              Skill Hub
+            
+              
+                Model Hub
+              
+              
+                Agent Hub
+              
+              
+                MCP Hub
+              
+              
+                Skill Hub
+              
             
 
             

               {/* Model Hub Tab */}
-              
+              
                 {/* Model Filters and Table */}
                 
                   {/* Header with Make Public Button */}
@@ -482,7 +490,7 @@ const ModelHubTable: React.FC = ({ accessToken, publicPage,
               
 
               {/* Agent Hub Tab */}
-              
+              
                 
                   {/* Header with Make Public Button */}
                   {publicPage == false && canModify && (
@@ -516,7 +524,7 @@ const ModelHubTable: React.FC = ({ accessToken, publicPage,
               
 
               {/* MCP Hub Tab */}
-              
+              
                 
                   {/* Header with Make Public Button */}
                   {publicPage == false && canModify && (
@@ -553,7 +561,7 @@ const ModelHubTable: React.FC = ({ accessToken, publicPage,
               
 
               {/* Skill Hub Tab */}
-              
+              
                 {publicPage == false && canModify && (
                   

                     
diff --git a/ui/litellm-dashboard/src/components/per_user_usage.tsx b/ui/litellm-dashboard/src/components/per_user_usage.tsx
index b239da83766..af45facb9a8 100644
--- a/ui/litellm-dashboard/src/components/per_user_usage.tsx
+++ b/ui/litellm-dashboard/src/components/per_user_usage.tsx
@@ -124,11 +124,11 @@ const PerUserUsage: React.FC = ({ accessToken, selectedTags,
       
Individual developer usage metrics

 
       
-        
-          
+        
+          
             User Details
           
-          
+          
             Usage Distribution
           
         
diff --git a/ui/litellm-dashboard/src/components/settings.tsx b/ui/litellm-dashboard/src/components/settings.tsx
index 2b9d8dad90b..a0bf3afdb20 100644
--- a/ui/litellm-dashboard/src/components/settings.tsx
+++ b/ui/litellm-dashboard/src/components/settings.tsx
@@ -491,7 +491,7 @@ const Settings: React.FC = ({ accessToken, userRole, userID,
             Alerting Settings
             Email Alerts
           
-          
+          
              = ({ accessToken, userRole, userID,
               }}
             />
           
-          
+          
             

               
             

           
-          
+          
             
               

                 Alerts are only supported for Slack Webhook URLs. Get your webhook urls from{" "}
@@ -601,10 +601,10 @@ const Settings: React.FC = ({ accessToken, userRole, userID,
               
             
           
-          
+          
             
           
-          
+          
             
           
         
diff --git a/ui/litellm-dashboard/src/components/templates/key_info_view.tsx b/ui/litellm-dashboard/src/components/templates/key_info_view.tsx
index c57a76a878a..e33ae8323f5 100644
--- a/ui/litellm-dashboard/src/components/templates/key_info_view.tsx
+++ b/ui/litellm-dashboard/src/components/templates/key_info_view.tsx
@@ -599,9 +599,13 @@ export default function KeyInfoView({
       
 
       
-        
-          Overview
-          Settings
+        
+          
+            Overview
+          
+          
+            Settings
+          
         
 
         

diff --git a/ui/litellm-dashboard/src/components/user_agent_activity.tsx b/ui/litellm-dashboard/src/components/user_agent_activity.tsx
index 74716fd866e..4a8b1d11204 100644
--- a/ui/litellm-dashboard/src/components/user_agent_activity.tsx
+++ b/ui/litellm-dashboard/src/components/user_agent_activity.tsx
@@ -485,11 +485,11 @@ const UserAgentActivity: React.FC = ({ accessToken, user
       
         
           
-            
-              
+            
+              
                 DAU/WAU/MAU
               
-              
+              
                 Per User Usage (Last 30 Days)
               
             
@@ -502,14 +502,14 @@ const UserAgentActivity: React.FC = ({ accessToken, user
               

 
               
-                
-                  
+                
+                  
                     DAU
                   
-                  
+                  
                     WAU
                   
-                  
+                  
                     MAU
                   
                 

From e126975468ee9dbc981b2458b4f7fed2c60c80f1 Mon Sep 17 00:00:00 2001
From: yuneng-jiang 
Date: Wed, 19 Aug 2026 13:32:00 -0700
Subject: [PATCH 5/9] refactor(ui): migrate the antd Button call sites onto the
 shadcn Button (#37505)

Moves all 58 antd Button JSX sites across 24 files onto the shadcn
Button, leaving zero antd Button importers.

Prop mapping follows what already merged rather than a new convention:
type="primary" to the default variant, a bare button to outline (the
house default), type="text" to ghost, type="link" to link,
type="dashed" to outline plus border-dashed, danger to destructive,
size="small" to sm, htmlType to type, block to w-full, the icon prop to
a child, and loading to disabled plus aria-busy. Icon-only buttons take
the matching icon-* size. Inline style props that had a direct utility
equivalent moved to className, and the opacity toggle on the create key
submit is dropped since the base cva already carries disabled:opacity-50.

Base UI's Button defaults type to "button" for native buttons, the same
default antd used, so bare buttons inside a form do not start
submitting.

guardrail_info.tsx and tag_info.tsx lose their last antd import, so
their no-restricted-imports suppressions are pruned. The other 22 files
keep other antd symbols and keep their entries.

One behavior change: the MCP transports docs link now opens in a new
tab, matching every other external docs link in the dashboard, instead
of navigating the dashboard away.

The ModelSettingsModal loading assertion moves off antd's spinner
element onto aria-busy, which is what the rest of the suite already
asserts; that markup cannot survive removing antd.
---
 ui/litellm-dashboard/eslint-suppressions.json |  6 ---
 .../_components/GuardrailConfig.tsx           | 21 +++++----
 .../guardrails/_components/guardrail_info.tsx | 37 ++++++++-------
 .../_components/EnvVarsSection.tsx            |  6 ++-
 .../_components/MCPPermissionManagement.tsx   |  6 ++-
 .../PassthroughAuthorizeSection.tsx           |  4 +-
 .../mcp-servers/_components/mcp_connect.tsx   | 25 ++++++----
 .../_components/mcp_server_edit.tsx           | 10 ++--
 .../prompts/_components/prompt_info.tsx       | 33 ++++++-------
 .../tag-management/_components/tag_info.tsx   | 12 ++---
 .../_components/view_users/user_info_view.tsx | 28 +++++------
 .../PluginSettings/PluginSettings.tsx         | 14 ++++--
 .../src/components/add_model/AddModelForm.tsx | 14 ++++--
 .../components/add_model/KeywordTierRules.tsx | 16 ++++---
 .../add_model/provider_specific_fields.tsx    |  8 +++-
 .../BudgetFallbacksEditor.tsx                 |  9 ++--
 .../key_team_helpers/BudgetWindowsEditor.tsx  | 13 ++++--
 .../components/model_add/CredentialModal.tsx  |  7 +--
 .../ModelSettingsModal.test.tsx               |  5 +-
 .../ModelSettingsModal/ModelSettingsModal.tsx | 10 ++--
 .../src/components/model_info_view.tsx        | 46 ++++++++++---------
 .../src/components/onboarding_link.tsx        |  7 ++-
 .../organisms/RegenerateKeyModal.tsx          | 17 +++++--
 .../organisms/create_key_button.tsx           | 10 ++--
 .../src/components/routing_groups/index.tsx   | 14 ++++--
 .../src/components/team/TeamInfo.tsx          | 35 +++++++-------
 26 files changed, 245 insertions(+), 168 deletions(-)

diff --git a/ui/litellm-dashboard/eslint-suppressions.json b/ui/litellm-dashboard/eslint-suppressions.json
index c471355db68..98225de5d8b 100644
--- a/ui/litellm-dashboard/eslint-suppressions.json
+++ b/ui/litellm-dashboard/eslint-suppressions.json
@@ -336,9 +336,6 @@
     "max-params": {
       "count": 1
     },
-    "no-restricted-imports": {
-      "count": 1
-    },
     "react-hooks/set-state-in-effect": {
       "count": 3
     }
@@ -1302,9 +1299,6 @@
     "local/filename-pascal-case": {
       "count": 1
     },
-    "no-restricted-imports": {
-      "count": 1
-    },
     "react-hooks/set-state-in-effect": {
       "count": 1
     }
diff --git a/ui/litellm-dashboard/src/app/(dashboard)/guardrails-monitor/_components/GuardrailConfig.tsx b/ui/litellm-dashboard/src/app/(dashboard)/guardrails-monitor/_components/GuardrailConfig.tsx
index 609af657f20..667c52087f1 100644
--- a/ui/litellm-dashboard/src/app/(dashboard)/guardrails-monitor/_components/GuardrailConfig.tsx
+++ b/ui/litellm-dashboard/src/app/(dashboard)/guardrails-monitor/_components/GuardrailConfig.tsx
@@ -5,7 +5,8 @@ import {
   RollbackOutlined,
   SaveOutlined,
 } from "@ant-design/icons";
-import { Button, Input, Select, Switch } from "antd";
+import { Input, Select, Switch } from "antd";
+import { Button } from "@/components/ui/button";
 import React, { useState } from "react";
 
 interface GuardrailConfigProps {
@@ -56,13 +57,17 @@ export function GuardrailConfig({ guardrailName, guardrailType, provider }: Guar
               options={versions.map((v) => ({ value: v.id, label: v.label }))}
               style={{ width: 140 }}
             />
-            
           

           

-            
-            
+            
           

@@ -194,12 +199,8 @@ export function GuardrailConfig({ guardrailName, guardrailType, provider }: Guar
         

 
         

-          
 
diff --git a/ui/litellm-dashboard/src/app/(dashboard)/guardrails/_components/guardrail_info.tsx b/ui/litellm-dashboard/src/app/(dashboard)/guardrails/_components/guardrail_info.tsx
index e80ddac932f..f89e8a3248c 100644
--- a/ui/litellm-dashboard/src/app/(dashboard)/guardrails/_components/guardrail_info.tsx
+++ b/ui/litellm-dashboard/src/app/(dashboard)/guardrails/_components/guardrail_info.tsx
@@ -5,10 +5,9 @@ import {
   updateGuardrailCall,
 } from "@/components/networking";
 import { copyToClipboard as utilCopyToClipboard } from "@/utils/dataUtils";
-import { CodeOutlined, EyeInvisibleOutlined, InfoCircleOutlined, StopOutlined } from "@ant-design/icons";
+import { EyeInvisibleOutlined, InfoCircleOutlined, StopOutlined } from "@ant-design/icons";
 
-import { Button as AntdButton } from "antd";
-import { ArrowLeft, CheckIcon, CopyIcon } from "lucide-react";
+import { ArrowLeft, CheckIcon, Code, CopyIcon } from "lucide-react";
 import { Badge } from "@/components/ui/badge";
 import { Card } from "@/components/ui/card";
 import { Tabs, TabsContent, TabsList, TabsTrigger } from "@/components/ui/tabs";
@@ -511,24 +510,26 @@ const GuardrailInfoView: React.FC = ({ guardrailId, onClose,
   return (
     

       

-        } onClick={onClose} className="mb-4">
+        
         
{guardrailData.guardrail_name || "Unnamed Guardrail"}

         

           
{guardrailData.guardrail_id}

 
-           : }
+          
         

       

 
@@ -628,13 +629,14 @@ const GuardrailInfoView: React.FC = ({ guardrailId, onClose,
               
                 

                   

-                    
+                    
                     
Custom Code

                   

                   {isAdmin && !isConfigGuardrail && (
-                    } onClick={() => setCustomCodeModalVisible(true)}>
+                    
                   )}
                 

                 

@@ -671,11 +673,14 @@ const GuardrailInfoView: React.FC = ({ guardrailId, onClose,
                   {!isEditing &&
                     !isConfigGuardrail &&
                     (guardrailData.litellm_params?.guardrail === "custom_code" ? (
-                      } onClick={() => setCustomCodeModalVisible(true)}>
+                      
                     ) : (
-                       setIsEditing(true)}>Edit Settings
+                      
                     ))}
                 

 
diff --git a/ui/litellm-dashboard/src/app/(dashboard)/mcp-servers/_components/EnvVarsSection.tsx b/ui/litellm-dashboard/src/app/(dashboard)/mcp-servers/_components/EnvVarsSection.tsx
index 539a06910c1..23a55d2dc1a 100644
--- a/ui/litellm-dashboard/src/app/(dashboard)/mcp-servers/_components/EnvVarsSection.tsx
+++ b/ui/litellm-dashboard/src/app/(dashboard)/mcp-servers/_components/EnvVarsSection.tsx
@@ -1,5 +1,6 @@
 import React from "react";
-import { Input, Select, Button, Tooltip, Typography } from "antd";
+import { Input, Select, Tooltip, Typography } from "antd";
+import { Button } from "@/components/ui/button";
 import { InfoCircleOutlined, MinusCircleOutlined, PlusOutlined } from "@ant-design/icons";
 import { useFieldArray, useFormContext, useWatch } from "react-hook-form";
 
@@ -106,7 +107,8 @@ const EnvVarsSection: React.FC = () => {
             

           

         ))}
-        
       

diff --git a/ui/litellm-dashboard/src/app/(dashboard)/mcp-servers/_components/MCPPermissionManagement.tsx b/ui/litellm-dashboard/src/app/(dashboard)/mcp-servers/_components/MCPPermissionManagement.tsx
index 3711140562f..b170157049c 100644
--- a/ui/litellm-dashboard/src/app/(dashboard)/mcp-servers/_components/MCPPermissionManagement.tsx
+++ b/ui/litellm-dashboard/src/app/(dashboard)/mcp-servers/_components/MCPPermissionManagement.tsx
@@ -1,5 +1,6 @@
 import React, { useEffect } from "react";
-import { Alert, Select, Tooltip, Collapse, Input, Space, Button, Switch } from "antd";
+import { Alert, Select, Tooltip, Collapse, Input, Space, Switch } from "antd";
+import { Button } from "@/components/ui/button";
 import { InfoCircleOutlined, MinusCircleOutlined, PlusOutlined } from "@ant-design/icons";
 import { useFieldArray, useFormContext, useWatch } from "react-hook-form";
 import { MCPServer, AUTH_TYPE } from "@/components/mcp_tools/types";
@@ -77,7 +78,8 @@ const StaticHeadersFieldArray: React.FC = () => {
           />
         
       ))}
-      
     
diff --git a/ui/litellm-dashboard/src/app/(dashboard)/mcp-servers/_components/PassthroughAuthorizeSection.tsx b/ui/litellm-dashboard/src/app/(dashboard)/mcp-servers/_components/PassthroughAuthorizeSection.tsx
index 375e1dfe515..406da95901a 100644
--- a/ui/litellm-dashboard/src/app/(dashboard)/mcp-servers/_components/PassthroughAuthorizeSection.tsx
+++ b/ui/litellm-dashboard/src/app/(dashboard)/mcp-servers/_components/PassthroughAuthorizeSection.tsx
@@ -1,5 +1,6 @@
 import React from "react";
-import { Button, Checkbox, Input } from "antd";
+import { Checkbox, Input } from "antd";
+import { Button } from "@/components/ui/button";
 import DcrBridgeToggle from "./DcrBridgeToggle";
 import { MountedFormField } from "@/components/common_components/MountedFormField";
 import { textControl } from "./mcpFieldRules";
@@ -119,6 +120,7 @@ export default function PassthroughAuthorizeSection({
         
       )}
       
         
{code}

       
     
@@ -441,11 +443,18 @@ const MCPConnect: React.FC = ({ currentServerAccessGroups = []
           />
           

             
           

diff --git a/ui/litellm-dashboard/src/app/(dashboard)/mcp-servers/_components/mcp_server_edit.tsx b/ui/litellm-dashboard/src/app/(dashboard)/mcp-servers/_components/mcp_server_edit.tsx
index 3892b9d37b7..eb462f788a5 100644
--- a/ui/litellm-dashboard/src/app/(dashboard)/mcp-servers/_components/mcp_server_edit.tsx
+++ b/ui/litellm-dashboard/src/app/(dashboard)/mcp-servers/_components/mcp_server_edit.tsx
@@ -1,5 +1,5 @@
 import React, { useState, useEffect } from "react";
-import { Select, Button as AntdButton, Tooltip, Input, InputNumber, Alert } from "antd";
+import { Select, Tooltip, Input, InputNumber, Alert } from "antd";
 import { FormProvider, useForm } from "react-hook-form";
 import { InfoCircleOutlined } from "@ant-design/icons";
 import { Button } from "@/components/ui/button";
@@ -1271,7 +1271,9 @@ const MCPServerEdit: React.FC = ({
                 
 
                 

-                  Cancel
+                  
                   
                 

               
@@ -1284,7 +1286,9 @@ const MCPServerEdit: React.FC = ({
             
 
             

-              Cancel
+              
               
             

           
diff --git a/ui/litellm-dashboard/src/app/(dashboard)/prompts/_components/prompt_info.tsx b/ui/litellm-dashboard/src/app/(dashboard)/prompts/_components/prompt_info.tsx
index cf29280ea51..5ce91f15bd4 100644
--- a/ui/litellm-dashboard/src/app/(dashboard)/prompts/_components/prompt_info.tsx
+++ b/ui/litellm-dashboard/src/app/(dashboard)/prompts/_components/prompt_info.tsx
@@ -4,7 +4,7 @@ import { Button } from "@/components/ui/button";
 import { Card } from "@/components/ui/card";
 import { Table, TableBody, TableCell, TableHead, TableHeader, TableRow } from "@/components/ui/table";
 import { Tabs, TabsContent, TabsList, TabsTrigger } from "@/components/ui/tabs";
-import { Button as AntButton, Modal } from "antd";
+import { Modal } from "antd";
 import {
   getPromptInfo,
   getPromptVersions,
@@ -191,17 +191,18 @@ const PromptInfoView: React.FC = ({ promptId, onClose, accessTo
             
Prompt Details

             

               
{basePromptId}

-               : }
+              
             

           
           

@@ -412,10 +413,9 @@ const PromptInfoView: React.FC = ({ promptId, onClose, accessTo
               
                 

                   
Prompt Template

-                   : }
+                  
                 

 
                 

@@ -462,10 +463,9 @@ const PromptInfoView: React.FC = ({ promptId, onClose, accessTo
             
               

                 
Raw API Response

-                 : }
+                
               

 
               

diff --git a/ui/litellm-dashboard/src/app/(dashboard)/tag-management/_components/tag_info.tsx b/ui/litellm-dashboard/src/app/(dashboard)/tag-management/_components/tag_info.tsx
index 5c561f38197..f05bddb3f6b 100644
--- a/ui/litellm-dashboard/src/app/(dashboard)/tag-management/_components/tag_info.tsx
+++ b/ui/litellm-dashboard/src/app/(dashboard)/tag-management/_components/tag_info.tsx
@@ -2,7 +2,6 @@
 
 import React, { useState, useEffect } from "react";
 import { SimpleTooltip } from "@/components/ui/tooltip";
-import { Button as AntdButton } from "antd";
 import { z } from "zod/v4";
 import { fetchUserModels } from "@/components/organisms/create_key_button";
 import { getModelDisplayName } from "@/components/key_team_helpers/fetch_available_models_team_key";
@@ -230,17 +229,18 @@ const TagInfoView: React.FC = ({ tagId, onClose, accessToken,
             
               {tagDetails.name}
             
-             : }
+            
           

           
{tagDetails.description || "No description"}

         

diff --git a/ui/litellm-dashboard/src/app/(dashboard)/users/_components/view_users/user_info_view.tsx b/ui/litellm-dashboard/src/app/(dashboard)/users/_components/view_users/user_info_view.tsx
index 5ea15df13f0..9fb4dad67e4 100644
--- a/ui/litellm-dashboard/src/app/(dashboard)/users/_components/view_users/user_info_view.tsx
+++ b/ui/litellm-dashboard/src/app/(dashboard)/users/_components/view_users/user_info_view.tsx
@@ -18,7 +18,7 @@ import {
   Member,
 } from "@/components/networking";
 import { SimpleTooltip } from "@/components/ui/tooltip";
-import { Button as AntdButton, Modal } from "antd";
+import { Modal } from "antd";
 import { Field, FieldGroup, FieldLabel } from "@/components/shared/form/field";
 import {
   Combobox,
@@ -405,17 +405,18 @@ export default function UserInfoView({
           
{userData.user_email || "User"}

           

             {userData.user_id}
-             : }
+            
           

         

         {userRole && rolesWithWriteAccess.includes(userRole) && (
@@ -585,17 +586,18 @@ export default function UserInfoView({
                   
User ID

                   

                     {userData.user_id}
-                     : }
+                    
                   

                 
 
@@ -760,9 +762,9 @@ export default function UserInfoView({
           
 
           

-            
+            
           

         
       
diff --git a/ui/litellm-dashboard/src/components/Settings/AdminSettings/PluginSettings/PluginSettings.tsx b/ui/litellm-dashboard/src/components/Settings/AdminSettings/PluginSettings/PluginSettings.tsx
index c93cf812738..fc085ebbd9e 100644
--- a/ui/litellm-dashboard/src/components/Settings/AdminSettings/PluginSettings/PluginSettings.tsx
+++ b/ui/litellm-dashboard/src/components/Settings/AdminSettings/PluginSettings/PluginSettings.tsx
@@ -1,7 +1,8 @@
 "use client";
 
 import { useState, useEffect } from "react";
-import { Button, Card, Modal, Space, Table, Typography } from "antd";
+import { Card, Modal, Space, Table, Typography } from "antd";
+import { Button } from "@/components/ui/button";
 import { DeleteOutlined, EditOutlined, PlusOutlined } from "@ant-design/icons";
 import { Eye, EyeOff } from "lucide-react";
 import { getConfigFieldSetting, updateConfigFieldSetting } from "@/components/networking";
@@ -113,8 +114,12 @@ export default function PluginSettings() {
       key: "actions",
       render: (_: unknown, __: Plugin, idx: number) => (
         
-          
+          
         
       ),
     },
@@ -131,7 +136,8 @@ export default function PluginSettings() {
         Each plugin must expose GET /api/plugin-manifest returning nav items and capabilities.
       
 
-      
 
diff --git a/ui/litellm-dashboard/src/components/add_model/AddModelForm.tsx b/ui/litellm-dashboard/src/components/add_model/AddModelForm.tsx
index 4d339737ab0..4cb60dc116a 100644
--- a/ui/litellm-dashboard/src/components/add_model/AddModelForm.tsx
+++ b/ui/litellm-dashboard/src/components/add_model/AddModelForm.tsx
@@ -5,7 +5,8 @@ import { all_admin_roles, isUserTeamAdminForAnyTeam } from "@/utils/roles";
 import { modelCreationScope } from "@/utils/modelPermissions";
 import { Switch } from "@/components/ui/switch";
 import { Field, FieldLabel } from "@/components/shared/form/field";
-import { Select as AntdSelect, Button, Card, Col, Modal, Row, Tooltip, Typography, Alert } from "antd";
+import { Select as AntdSelect, Card, Col, Modal, Row, Tooltip, Typography, Alert } from "antd";
+import { Button } from "@/components/ui/button";
 import type { UploadProps } from "antd/es/upload";
 import React, { useEffect, useMemo, useState } from "react";
 import { FormProvider, useWatch, type UseFormReturn } from "react-hook-form";
@@ -418,10 +419,16 @@ const AddModelForm: React.FC = ({
                     Need Help?
                   
                   

-                    
-                    
                   

@@ -443,6 +450,7 @@ const AddModelForm: React.FC = ({
         footer={[
           
       
@@ -131,12 +133,14 @@ const KeywordTierRules: React.FC = ({ rules, onChange, ti
                   />
                 
                 
               
             
           ))}
diff --git a/ui/litellm-dashboard/src/components/add_model/provider_specific_fields.tsx b/ui/litellm-dashboard/src/components/add_model/provider_specific_fields.tsx
index 955167c4351..e5f1f108b4f 100644
--- a/ui/litellm-dashboard/src/components/add_model/provider_specific_fields.tsx
+++ b/ui/litellm-dashboard/src/components/add_model/provider_specific_fields.tsx
@@ -1,7 +1,8 @@
 import { useProviderFields } from "@/app/(dashboard)/hooks/providers/useProviderFields";
 import { UploadOutlined } from "@ant-design/icons";
 import { Input } from "@/components/ui/input";
-import { Button as Button2, Col, Input as AntdInput, Row, Select, Typography, Upload, UploadProps } from "antd";
+import { Col, Input as AntdInput, Row, Select, Typography, Upload, UploadProps } from "antd";
+import { Button } from "@/components/ui/button";
 import React from "react";
 import { useFormContext } from "react-hook-form";
 import { antdRequired } from "../common_components/antdFormRules";
@@ -255,7 +256,10 @@ const ProviderSpecificFields: React.FC = ({ selecte
             }
           }}
         >
-          }>Click to Upload
+          
         
       );
     }
diff --git a/ui/litellm-dashboard/src/components/key_team_helpers/BudgetFallbacksEditor.tsx b/ui/litellm-dashboard/src/components/key_team_helpers/BudgetFallbacksEditor.tsx
index 062aaaa1773..1d327b87346 100644
--- a/ui/litellm-dashboard/src/components/key_team_helpers/BudgetFallbacksEditor.tsx
+++ b/ui/litellm-dashboard/src/components/key_team_helpers/BudgetFallbacksEditor.tsx
@@ -1,5 +1,6 @@
 import { SimpleTooltip } from "@/components/ui/tooltip";
-import { Button, Select } from "antd";
+import { Select } from "antd";
+import { Button } from "@/components/ui/button";
 import { ArrowDown, Plus, X } from "lucide-react";
 import React, { useState } from "react";
 
@@ -62,7 +63,8 @@ export function BudgetFallbacksEditor({ value, onChange, availableModels }: Budg
         

           When a model exceeds its per-model budget, requests automatically reroute to fallback models
         

-        
       
@@ -143,7 +145,8 @@ export function BudgetFallbacksEditor({ value, onChange, availableModels }: Budg
           
         );
       })}
-      
     
diff --git a/ui/litellm-dashboard/src/components/key_team_helpers/BudgetWindowsEditor.tsx b/ui/litellm-dashboard/src/components/key_team_helpers/BudgetWindowsEditor.tsx
index 0c90674a479..c9a422dcb85 100644
--- a/ui/litellm-dashboard/src/components/key_team_helpers/BudgetWindowsEditor.tsx
+++ b/ui/litellm-dashboard/src/components/key_team_helpers/BudgetWindowsEditor.tsx
@@ -1,4 +1,5 @@
-import { Button, InputNumber, Select } from "antd";
+import { InputNumber, Select } from "antd";
+import { Button } from "@/components/ui/button";
 import React from "react";
 
 export interface BudgetWindowEntry {
@@ -55,7 +56,12 @@ export function BudgetWindowsEditor({ value, onChange }: BudgetWindowsEditorProp
                 style={{ width: 160 }}
                 prefix="$"
               />
-              
             
@@ -64,7 +70,8 @@ export function BudgetWindowsEditor({ value, onChange }: BudgetWindowsEditorProp
         );
       })}
       
-                
+                
               
             
           
diff --git a/ui/litellm-dashboard/src/components/model_dashboard/ModelSettingsModal/ModelSettingsModal.test.tsx b/ui/litellm-dashboard/src/components/model_dashboard/ModelSettingsModal/ModelSettingsModal.test.tsx
index fb6624ed92d..ccf5bb062b2 100644
--- a/ui/litellm-dashboard/src/components/model_dashboard/ModelSettingsModal/ModelSettingsModal.test.tsx
+++ b/ui/litellm-dashboard/src/components/model_dashboard/ModelSettingsModal/ModelSettingsModal.test.tsx
@@ -2,7 +2,7 @@ import { useProxyConfig } from "@/app/(dashboard)/hooks/proxyConfig/useProxyConf
 import { useStoreModelInDB } from "@/app/(dashboard)/hooks/storeModelInDB/useStoreModelInDB";
 import { toast } from "@/lib/toast";
 import { parseErrorMessage } from "@/components/shared/errorUtils";
-import { screen, waitFor, within } from "@testing-library/react";
+import { screen, waitFor } from "@testing-library/react";
 import userEvent from "@testing-library/user-event";
 import { beforeEach, describe, expect, it, vi } from "vitest";
 import { renderWithProviders } from "../../../../tests/test-utils";
@@ -225,7 +225,8 @@ describe("ModelSettingsModal", () => {
 
     const saveButton = screen.getByRole("button", { name: /Saving/i });
     expect(saveButton).toBeInTheDocument();
-    expect(within(saveButton).getByRole("img", { name: "loading" })).toBeInTheDocument();
+    expect(saveButton).toHaveAttribute("aria-busy", "true");
+    expect(saveButton).toBeDisabled();
   });
 
   it("should not render modal when isVisible is false", () => {
diff --git a/ui/litellm-dashboard/src/components/model_dashboard/ModelSettingsModal/ModelSettingsModal.tsx b/ui/litellm-dashboard/src/components/model_dashboard/ModelSettingsModal/ModelSettingsModal.tsx
index 01696e2d820..400b1c2859d 100644
--- a/ui/litellm-dashboard/src/components/model_dashboard/ModelSettingsModal/ModelSettingsModal.tsx
+++ b/ui/litellm-dashboard/src/components/model_dashboard/ModelSettingsModal/ModelSettingsModal.tsx
@@ -8,7 +8,8 @@ import { FieldGroup } from "@/components/shared/form/field";
 import { FormField } from "@/components/shared/form/FormField";
 import { Switch } from "@/components/ui/switch";
 import { Tooltip, TooltipContent, TooltipProvider, TooltipTrigger } from "@/components/ui/tooltip";
-import { Button, Modal, Skeleton, Space, Typography } from "antd";
+import { Modal, Skeleton, Space, Typography } from "antd";
+import { Button } from "@/components/ui/button";
 import { CircleHelp } from "lucide-react";
 import React, { useEffect, useMemo } from "react";
 import { useForm } from "react-hook-form";
@@ -85,13 +86,12 @@ const ModelSettingsModal: React.FC = ({ isVisible, onCa
       open={isVisible}
       footer={
         
-          
           
           
         
         

           {(!isAnyAutoRouter || isComplexityRouterModel) && (
-            }
+            
           )}
 
           {!isAnyAutoRouter && (
             <>
-              }
+              
 
-              }
+              
             
           )}
-          }
+          
         

       
 
@@ -865,9 +869,9 @@ export default function ModelInfoView({
         open={isAutoRouterTestModalOpen}
         onCancel={() => setIsAutoRouterTestModalOpen(false)}
         footer={[
-           setIsAutoRouterTestModalOpen(false)}>
+          ,
         ]}
         width={700}
       >
diff --git a/ui/litellm-dashboard/src/components/onboarding_link.tsx b/ui/litellm-dashboard/src/components/onboarding_link.tsx
index 82c99119cf4..583a0d25b29 100644
--- a/ui/litellm-dashboard/src/components/onboarding_link.tsx
+++ b/ui/litellm-dashboard/src/components/onboarding_link.tsx
@@ -1,5 +1,6 @@
 import React from "react";
-import { Button, Modal, Typography } from "antd";
+import { Modal, Typography } from "antd";
+import { Button } from "@/components/ui/button";
 import { CopyToClipboard } from "react-copy-to-clipboard";
 import { toast } from "@/lib/toast";
 
@@ -98,9 +99,7 @@ export default function OnboardingModal({
       
       

          toast.success("Copied!")}>
-          
+          
         
       

     
diff --git a/ui/litellm-dashboard/src/components/organisms/RegenerateKeyModal.tsx b/ui/litellm-dashboard/src/components/organisms/RegenerateKeyModal.tsx
index 2edd502eee3..d717c11de86 100644
--- a/ui/litellm-dashboard/src/components/organisms/RegenerateKeyModal.tsx
+++ b/ui/litellm-dashboard/src/components/organisms/RegenerateKeyModal.tsx
@@ -1,6 +1,7 @@
 import useAuthorized from "@/app/(dashboard)/hooks/useAuthorized";
 import { CheckOutlined, CopyOutlined, SyncOutlined } from "@ant-design/icons";
-import { Alert, Button, Modal, Space } from "antd";
+import { Alert, Modal, Space } from "antd";
+import { Button } from "@/components/ui/button";
 import { CircleHelp } from "lucide-react";
 import React, { useEffect, useMemo, useState } from "react";
 import { useWatch } from "react-hook-form";
@@ -157,9 +158,12 @@ export function RegenerateKeyModal({ selectedToken, visible, onClose, onKeyUpdat
         regeneratedKey
           ? [
               
-                
+                
                 
-                  
                 
@@ -167,8 +171,11 @@ export function RegenerateKeyModal({ selectedToken, visible, onClose, onKeyUpdat
             ]
           : [
               
-                
-                
+                
               ,
diff --git a/ui/litellm-dashboard/src/components/organisms/create_key_button.tsx b/ui/litellm-dashboard/src/components/organisms/create_key_button.tsx
index ad7869202dc..0c78319bd6e 100644
--- a/ui/litellm-dashboard/src/components/organisms/create_key_button.tsx
+++ b/ui/litellm-dashboard/src/components/organisms/create_key_button.tsx
@@ -13,7 +13,7 @@ import { Button } from "@/components/ui/button";
 import { Collapsible, CollapsibleContent, CollapsibleTrigger } from "@/components/ui/collapsible";
 import { Input } from "@/components/ui/input";
 import { Field, FieldLabel } from "@/components/shared/form/field";
-import { Button as Button2, Input as AntdInput, Modal, Radio, Select, Switch, Tag, Tooltip, Typography } from "antd";
+import { Input as AntdInput, Modal, Radio, Select, Switch, Tag, Tooltip, Typography } from "antd";
 import { ChevronDown } from "lucide-react";
 import { useDebouncedCallback } from "@tanstack/react-pacer/debouncer";
 import { DEBOUNCE_WAIT_MS } from "@/utils/debounceConstants";
@@ -708,9 +708,9 @@ const CreateKey: React.FC = ({ team, teams, data, addKey, autoOp
                           style={{ width: "100%" }}
                           notFoundContent={userSearchLoading ? "Searching..." : "No users found"}
                         />
-                         setIsCreateUserModalVisible(true)} style={{ marginLeft: "8px" }}>
+                        
                       
                       
Search by email to find users

                     
@@ -1726,9 +1726,9 @@ const CreateKey: React.FC = ({ team, teams, data, addKey, autoOp
             )}
 
             

-              
+              
             

           
         
diff --git a/ui/litellm-dashboard/src/components/routing_groups/index.tsx b/ui/litellm-dashboard/src/components/routing_groups/index.tsx
index 41e71c77e1f..1ecea5b10da 100644
--- a/ui/litellm-dashboard/src/components/routing_groups/index.tsx
+++ b/ui/litellm-dashboard/src/components/routing_groups/index.tsx
@@ -1,7 +1,8 @@
 "use client";
 
 import React, { useMemo, useState } from "react";
-import { Button, Card, Flex, Input, Modal, Space, Typography } from "antd";
+import { Card, Flex, Input, Modal, Space, Typography } from "antd";
+import { Button } from "@/components/ui/button";
 import { PlusOutlined, ReloadOutlined, SearchOutlined } from "@ant-design/icons";
 import { useRoutingGroups, useSaveRoutingGroups } from "@/app/(dashboard)/hooks/routingGroups/useRoutingGroups";
 import { useRouterFields } from "@/app/(dashboard)/hooks/router/useRouterFields";
@@ -112,10 +113,17 @@ const RoutingGroups: React.FC = () => {
             className="max-w-sm"
           />
           
-            
-            
             
diff --git a/ui/litellm-dashboard/src/components/team/TeamInfo.tsx b/ui/litellm-dashboard/src/components/team/TeamInfo.tsx
index c20f47a155b..1684c3b1225 100644
--- a/ui/litellm-dashboard/src/components/team/TeamInfo.tsx
+++ b/ui/litellm-dashboard/src/components/team/TeamInfo.tsx
@@ -27,7 +27,7 @@ import { Badge } from "@/components/ui/badge";
 import { Card } from "@/components/ui/card";
 import { Collapsible, CollapsibleContent, CollapsibleTrigger } from "@/components/ui/collapsible";
 import { Input as UIInput } from "@/components/ui/input";
-import { Button as UIButton } from "@/components/ui/button";
+import { Button } from "@/components/ui/button";
 import { Switch } from "@/components/ui/switch";
 import { Textarea } from "@/components/ui/textarea";
 import { TooltipProvider } from "@/components/ui/tooltip";
@@ -39,7 +39,7 @@ import { MultiSelect } from "@/components/shared/MultiSelect";
 import { SearchSelect } from "@/components/shared/SearchSelect";
 import { useZodForm } from "@/lib/forms/useZodForm";
 import { TagsInput } from "@/app/(dashboard)/guardrails/_components/content_filter/TagsInput";
-import { Button, Tabs, Tooltip } from "antd";
+import { Tabs, Tooltip } from "antd";
 import { toast } from "@/lib/toast";
 import { CheckIcon, ChevronDown, CircleMinus, CopyIcon, Plus, Save } from "lucide-react";
 import React, { useEffect, useMemo, useState } from "react";
@@ -950,23 +950,25 @@ const TeamInfoView: React.FC = ({
     

       

         

-          
           
{info.team_alias}

           

             
{info.team_id}

             
           

         

       

@@ -1149,12 +1151,13 @@ const TeamInfoView: React.FC = ({
                   
Team Settings

                   {canEditTeam && !isEditing && (
                     
                   )}
@@ -1444,7 +1447,7 @@ const TeamInfoView: React.FC = ({
                                   />
                                 )}
                               
-                               = ({
                                 onClick={() => removeModelLimit(index)}
                               >
                                 
-                              
+                              
                             

                           ))}
-                           = ({
                           >
                             
                             Add Model Limit
-                          
+                          
                         
 
                          = ({
 
                       

                         

-                           setIsEditing(false)}
                             disabled={isTeamSaving}
                           >
                             Cancel
-                          
-                          
+                          
+                          
                         

                       

                     

From 7675ba8717c920d9699ce8eda7a6666789e82dcd Mon Sep 17 00:00:00 2001
From: yuneng-jiang 
Date: Wed, 19 Aug 2026 13:40:14 -0700
Subject: [PATCH 6/9] test(ui): split the vitest suite into unit, component,
 integration and type projects (#37488)

* test(ui): split the vitest suite into unit, component and integration tiers

Every test file booted jsdom, including the ~1800 that assert pure functions
and never render. They now run as a separate vitest project in the node
environment, where the whole tier finishes in under four seconds.

The tiers are vitest projects rather than a naming convention, so CI can run
them as independent jobs. A .test.ts that renders React, a hook test being the
usual case, is listed explicitly and stays in the jsdom tier.

* test(ui): report per-test duration against a per-tier budget

A timeout only catches a hung test, and it has to stay generous enough to
survive a loaded runner, so it never reports the multi-second render tests that
make CI fail the moment the box is busy. Budgets are separate and far tighter:
50ms unit, 1s component, 3s integration.

The counts are laptop measurements, so the CI job is report-only for now.
Flipping it to blocking is one line once CI has published its own numbers.

* test(ui): run the tiers as separate CI jobs and stop clicking popups by text

The old job ran every file in one process, so the single slowest file set the
wall clock and a bigger box bought nothing. The tiers now run as separate jobs
with the component tier sharded four ways.

getByText and findByText match hidden nodes, so they resolve against a closed
Base UI popup whose positioner still carries pointer-events: none, and the
click lands or not depending on how far the open transition got. Two files
failed this way, one three runs in five and one every run. Querying the option
by role waits for it to be visible, and both are now stable. A lint rule keeps
the pattern from coming back.

* test(ui): give React Testing Library's async queries a CI-sized window

findBy* and waitFor run on asyncUtilTimeout, which defaults to 1000ms and is
independent of vitest's testTimeout. Raising the vitest timeout therefore did
nothing for them: a query still gave up after one second while the test had 59
seconds of budget left, which is why a loaded runner produced 'Unable to find
role=...' rather than a timeout.

UserSearchModal is the worked example. The role query it makes resolves in
249ms on a laptop and blew past 1000ms on CI, failing the run at 1494ms. Five
seconds keeps the same assertions and only widens the window a failing query
waits before reporting; a passing query still resolves the moment the element
appears.

* test(ui): calibrate the tier budgets from real CI numbers and report by default

The first CI run showed the laptop counts were badly off: component 176 local
against 326 on CI, integration 87 against 128. The maxima now come from that
run with headroom.

continue-on-error still painted the check red, which is the opposite of the
point, so the report-only decision moves into test-budgets.json as an explicit
enforce flag. The job passes and prints the counts; flipping enforce to true
makes it a gate.

* docs(ui): drop the CLAUDE.md edits from the tier split

Keeping this PR to the vitest, CI and test changes.

* ci(ui): run every tier in one job instead of eight check rows

Sharding bought nothing. Measured on the first run of this branch, the
component tier unsharded finishes in 198s while the integration tier is floored
at 384s by a single file, so integration was always the critical path and the
four component shards only added rows. One job running every project comes in
around 384s against the 426s the split jobs took.

Eight rows named things like 'component (2)' also told a reviewer nothing, on a
PR page that already carries forty checks.

The job keeps the id ui-unit-tests because guard-internal-staging requires that
exact context; renaming the jobs had silently stopped it reporting, which would
have blocked every merge on a check that no longer existed. The workflow's
display name becomes UI Tests since it runs more than unit tests.

The tier split itself is untouched: it lives in the vitest projects config, so
the unit tier still runs in node with no jsdom, and each tier keeps its own
timeout and budget.

* fix(ui): stop the type check from running the whole suite a second time

test:types was 'vitest --run --typecheck.only'. Under test.projects that flag
is ignored and the root-level typecheck block is not inherited, so the step
collected each project's normal include and ran all 8464 runtime tests instead
of type-checking. It took 542s on CI against 33s on the flat config it
replaced, and the job then ran the same suite again in the next step.

Typecheck now belongs to a project of its own, with an empty include so it
contributes no runtime tests, and the CI job runs one vitest invocation for all
four. The type tier adds about 3s to a full run and reports 'Type Errors: no
errors' rather than a suite of tests.

Verified it still catches things: breaking SortingState in DataTable.test-d.tsx
fails with 'Type number is not assignable to type string' and exit 1, and
restoring it passes.

* test(ui): scope the split down to the vitest tier projects

Removes everything from this branch that was not the tier split.

The three lint rules brought 1381 lines of grandfathered suppressions in
eslint-suppressions.json, which is 81% of the branch's added lines and
debt nobody is going to pay down. The per-test duration budget does not
scale as a CI step. Both are gone, along with the two query rewrites the
no-click-by-text rule forced: those files pass 10/10 at this base, quiet
and under load, so there was no failure behind them.

The workflow is byte-identical to the base again. It already runs
npm run test:types and then vitest related on pull requests, so PR cost
is unchanged; the split only repoints test:types at the new project.
That project is required, not optional: vitest silently ignores
--typecheck.only under test.projects, so without it the type script
collects the whole suite instead of the one typed file.

Restores the base 60s testTimeout on the unit tier. The 5s cap was not
part of the split and failed ChatShell.serverRootPath.test.ts, a 960ms
test, under load.
---
 ui/litellm-dashboard/package.json             |   5 +-
 .../multi_export_utils.test.ts                |   2 +
 .../_components/createOAuthUiState.test.ts    |   2 +
 .../EntityUsageExport/utils.test.ts           |   2 +
 .../src/utils/cookieUtils.test.ts             |   2 +
 .../src/utils/dataUtils.test.ts               |   2 +
 .../src/utils/localStorageUtils.test.ts       |   2 +
 .../src/utils/mcpTokenStore.test.ts           |   2 +
 .../src/utils/returnUrlUtils.test.ts          |   2 +
 ui/litellm-dashboard/tests/setup.unit.ts      |  12 +++
 ui/litellm-dashboard/vitest.config.ts         | 101 +++++++++++++-----
 11 files changed, 104 insertions(+), 30 deletions(-)
 create mode 100644 ui/litellm-dashboard/tests/setup.unit.ts

diff --git a/ui/litellm-dashboard/package.json b/ui/litellm-dashboard/package.json
index 9407ff25a1f..ac13a12620d 100644
--- a/ui/litellm-dashboard/package.json
+++ b/ui/litellm-dashboard/package.json
@@ -9,8 +9,11 @@
     "start": "next start",
     "lint": "eslint .",
     "test": "vitest",
+    "test:unit": "vitest run --project unit",
+    "test:component": "vitest run --project component",
+    "test:integration": "vitest run --project integration",
     "test:dot": "vitest --reporter=dot",
-    "test:types": "vitest --run --typecheck.only",
+    "test:types": "vitest run --project types",
     "test:watch": "vitest -w",
     "test:coverage": "vitest run --coverage",
     "format": "prettier --write .",
diff --git a/ui/litellm-dashboard/src/app/(dashboard)/cost-tracking/_components/pricing_calculator/multi_export_utils.test.ts b/ui/litellm-dashboard/src/app/(dashboard)/cost-tracking/_components/pricing_calculator/multi_export_utils.test.ts
index 3c319e5ba18..25b4da869bb 100644
--- a/ui/litellm-dashboard/src/app/(dashboard)/cost-tracking/_components/pricing_calculator/multi_export_utils.test.ts
+++ b/ui/litellm-dashboard/src/app/(dashboard)/cost-tracking/_components/pricing_calculator/multi_export_utils.test.ts
@@ -1,3 +1,5 @@
+// @vitest-environment jsdom
+
 import { describe, it, expect, vi, beforeEach, afterEach } from "vitest";
 import { exportMultiToPDF, exportMultiToCSV } from "./multi_export_utils";
 import type { MultiModelResult } from "./types";
diff --git a/ui/litellm-dashboard/src/app/(dashboard)/mcp-servers/_components/createOAuthUiState.test.ts b/ui/litellm-dashboard/src/app/(dashboard)/mcp-servers/_components/createOAuthUiState.test.ts
index e2e7814f0d7..94544ec6e2d 100644
--- a/ui/litellm-dashboard/src/app/(dashboard)/mcp-servers/_components/createOAuthUiState.test.ts
+++ b/ui/litellm-dashboard/src/app/(dashboard)/mcp-servers/_components/createOAuthUiState.test.ts
@@ -1,3 +1,5 @@
+// @vitest-environment jsdom
+
 import { beforeEach, describe, expect, it, vi } from "vitest";
 import { setSecureItem } from "@/utils/secureStorage";
 import { CreateUiSnapshot, readCreateUiSnapshot, writeCreateUiSnapshot } from "./createOAuthUiState";
diff --git a/ui/litellm-dashboard/src/components/EntityUsageExport/utils.test.ts b/ui/litellm-dashboard/src/components/EntityUsageExport/utils.test.ts
index d80294aa2e6..7ed014d43b2 100644
--- a/ui/litellm-dashboard/src/components/EntityUsageExport/utils.test.ts
+++ b/ui/litellm-dashboard/src/components/EntityUsageExport/utils.test.ts
@@ -1,3 +1,5 @@
+// @vitest-environment jsdom
+
 import type { DateRangePickerValue } from "@/components/shared/date_picker_types";
 import Papa from "papaparse";
 import { afterEach, beforeEach, describe, expect, it, vi } from "vitest";
diff --git a/ui/litellm-dashboard/src/utils/cookieUtils.test.ts b/ui/litellm-dashboard/src/utils/cookieUtils.test.ts
index cc4fe64e4c8..a38b96e74c2 100644
--- a/ui/litellm-dashboard/src/utils/cookieUtils.test.ts
+++ b/ui/litellm-dashboard/src/utils/cookieUtils.test.ts
@@ -1,3 +1,5 @@
+// @vitest-environment jsdom
+
 import { describe, it, expect, beforeEach, vi } from "vitest";
 import { clearTokenCookies, getCookie, storeLoginToken } from "./cookieUtils";
 import { getToken, setToken } from "./mcpTokenStore";
diff --git a/ui/litellm-dashboard/src/utils/dataUtils.test.ts b/ui/litellm-dashboard/src/utils/dataUtils.test.ts
index d975a4a73f4..f696780bc0a 100644
--- a/ui/litellm-dashboard/src/utils/dataUtils.test.ts
+++ b/ui/litellm-dashboard/src/utils/dataUtils.test.ts
@@ -1,3 +1,5 @@
+// @vitest-environment jsdom
+
 import { describe, it, expect, beforeEach, vi, afterEach } from "vitest";
 import { copyToClipboard, formatNumberWithCommas, getSpendString, updateExistingKeys } from "./dataUtils";
 
diff --git a/ui/litellm-dashboard/src/utils/localStorageUtils.test.ts b/ui/litellm-dashboard/src/utils/localStorageUtils.test.ts
index 934c358cb74..2ef5e1d5b43 100644
--- a/ui/litellm-dashboard/src/utils/localStorageUtils.test.ts
+++ b/ui/litellm-dashboard/src/utils/localStorageUtils.test.ts
@@ -1,3 +1,5 @@
+// @vitest-environment jsdom
+
 import { describe, it, expect, vi, beforeEach } from "vitest";
 import {
   emitLocalStorageChange,
diff --git a/ui/litellm-dashboard/src/utils/mcpTokenStore.test.ts b/ui/litellm-dashboard/src/utils/mcpTokenStore.test.ts
index 0d096677f0d..e384f41042a 100644
--- a/ui/litellm-dashboard/src/utils/mcpTokenStore.test.ts
+++ b/ui/litellm-dashboard/src/utils/mcpTokenStore.test.ts
@@ -1,3 +1,5 @@
+// @vitest-environment jsdom
+
 import { afterEach, beforeEach, describe, expect, it } from "vitest";
 import { clearAllMcpTokens, getToken, isTokenValid, removeToken, setToken } from "./mcpTokenStore";
 
diff --git a/ui/litellm-dashboard/src/utils/returnUrlUtils.test.ts b/ui/litellm-dashboard/src/utils/returnUrlUtils.test.ts
index 0f3a8b4bf69..3d1e83ad236 100644
--- a/ui/litellm-dashboard/src/utils/returnUrlUtils.test.ts
+++ b/ui/litellm-dashboard/src/utils/returnUrlUtils.test.ts
@@ -1,3 +1,5 @@
+// @vitest-environment jsdom
+
 import {
   buildLoginUrlWithReturn,
   clearStoredReturnUrl,
diff --git a/ui/litellm-dashboard/tests/setup.unit.ts b/ui/litellm-dashboard/tests/setup.unit.ts
new file mode 100644
index 00000000000..4eb0c473989
--- /dev/null
+++ b/ui/litellm-dashboard/tests/setup.unit.ts
@@ -0,0 +1,12 @@
+import { vi } from "vitest";
+
+vi.mock("@/lib/toast", () => ({
+  toast: {
+    success: vi.fn(),
+    info: vi.fn(),
+    warning: vi.fn(),
+    error: vi.fn(),
+    fromError: vi.fn(),
+    dismiss: vi.fn(),
+  },
+}));
diff --git a/ui/litellm-dashboard/vitest.config.ts b/ui/litellm-dashboard/vitest.config.ts
index 7d8825909d8..ca2a25cba9f 100644
--- a/ui/litellm-dashboard/vitest.config.ts
+++ b/ui/litellm-dashboard/vitest.config.ts
@@ -12,17 +12,81 @@ const staticImageData: Plugin = {
   },
 };
 
-const config: ViteUserConfig = {
+const sharedViteConfig = {
   plugins: [staticImageData],
+  resolve: { alias: { "@": resolve(__dirname, "src") } },
+  define: { "import.meta.vitest": "undefined" },
+  esbuild: { jsx: "automatic", jsxImportSource: "react" } as const,
+};
+
+const TEST_TS_FILES_THAT_RENDER_REACT: readonly string[] = [
+  "src/**/hooks/**/*.test.ts",
+  "src/**/cost-tracking/_components/**/use_*.test.ts",
+  "src/**/models-and-endpoints/detailNavigation.test.ts",
+  "src/**/models-and-endpoints/vertexCredentialsUpload.test.ts",
+  "src/components/chat/useChatHistory.test.ts",
+  "src/lib/forms/pickDirty.test.ts",
+];
+
+const jsdomTier = {
+  environment: "./tests/jsdomFetchEnv.ts",
+  setupFiles: ["tests/setupTests.ts"],
+  globals: true,
+  css: true,
+  testTimeout: 60_000,
+  hookTimeout: 30_000,
+};
+
+const config: ViteUserConfig = {
+  ...sharedViteConfig,
   test: {
-    environment: "./tests/jsdomFetchEnv.ts",
-    setupFiles: ["tests/setupTests.ts"],
-    globals: true,
-    css: true, // lets you import CSS/modules without extra mocks
-    testTimeout: 60000,
-    hookTimeout: 30000,
+    projects: [
+      {
+        ...sharedViteConfig,
+        test: {
+          name: "unit",
+          environment: "node",
+          setupFiles: ["tests/setup.unit.ts"],
+          globals: true,
+          testTimeout: 60_000,
+          hookTimeout: 30_000,
+          include: ["src/**/*.test.ts", "tests/**/*.test.ts"],
+          exclude: ["node_modules/**", ...TEST_TS_FILES_THAT_RENDER_REACT],
+        },
+      },
+      {
+        ...sharedViteConfig,
+        test: {
+          ...jsdomTier,
+          name: "component",
+          include: ["src/**/*.test.tsx", "tests/**/*.test.tsx", ...TEST_TS_FILES_THAT_RENDER_REACT],
+          exclude: ["node_modules/**", "**/*.integration.test.tsx"],
+        },
+      },
+      {
+        ...sharedViteConfig,
+        test: {
+          ...jsdomTier,
+          name: "integration",
+          include: ["src/**/*.integration.test.tsx", "tests/**/*.integration.test.tsx"],
+          exclude: ["node_modules/**"],
+        },
+      },
+      {
+        ...sharedViteConfig,
+        test: {
+          name: "types",
+          include: [],
+          typecheck: {
+            enabled: true,
+            include: ["src/**/*.test-d.ts", "src/**/*.test-d.tsx"],
+            ignoreSourceErrors: true,
+          },
+        },
+      },
+    ],
     silent: process.env.CI ? "passed-only" : false,
-    teardownTimeout: 60000,
+    teardownTimeout: 60_000,
     coverage: {
       provider: "v8",
       reporter: ["text", "lcov"],
@@ -32,37 +96,16 @@ const config: ViteUserConfig = {
         "**/*.test.*",
         "**/*.test-d.*",
         "**/*.spec.*",
-
         "tests/**",
-
         "node_modules/**",
         ".next/**",
         "out/**",
-
         "**/*.config.*",
         "postcss.config.*",
         "tailwind.config.*",
         "next.config.*",
       ],
     },
-    exclude: ["node_modules/**"],
-    include: ["src/**/*.test.ts", "src/**/*.test.tsx", "tests/**/*.test.ts", "tests/**/*.test.tsx"],
-    typecheck: {
-      include: ["src/**/*.test-d.ts", "src/**/*.test-d.tsx"],
-      ignoreSourceErrors: true,
-    },
-  },
-  resolve: {
-    alias: {
-      "@": resolve(__dirname, "src"),
-    },
-  },
-  define: {
-    "import.meta.vitest": "undefined",
-  },
-  esbuild: {
-    jsx: "automatic",
-    jsxImportSource: "react",
   },
 };
 

From a1afc2f433edf0d4961512713cec94761d112c54 Mon Sep 17 00:00:00 2001
From: yucheng-berri 
Date: Wed, 19 Aug 2026 13:51:31 -0700
Subject: [PATCH 7/9] refactor(ptu): give the rollup a source-agnostic
 deployment record (#37501)

The flat-cost rollup reads deployments only from LiteLLM_ProxyModelTable, so a PTU
deployment declared in config.yaml never accrues flat cost. Those deployments live in
llm_router.model_list as plain dicts whose id sits in model_info rather than on the entry,
so they do not satisfy the shape _parse_ptu_model reads.

Adds a frozen record in that shape and a factory that maps a router entry onto it, leaving
_parse_ptu_model byte-identical so the existing cases stand as evidence of no behaviour
change. Nothing calls the factory yet; the caller lands with the loader union.

_decode_model_info also stops handing back valid JSON that is not an object. It decoded
a list or a scalar and returned it as a mapping, so the caller read fields off it and
raised, losing the whole run rather than the one bad deployment.
---
 .../spend_tracking/ptu_flat_cost_rollup.py    |  46 ++++-
 .../test_ptu_flat_cost_rollup.py              | 177 ++++++++++++++++++
 2 files changed, 220 insertions(+), 3 deletions(-)

diff --git a/litellm/proxy/spend_tracking/ptu_flat_cost_rollup.py b/litellm/proxy/spend_tracking/ptu_flat_cost_rollup.py
index efdbda47fdc..25de9f6d065 100644
--- a/litellm/proxy/spend_tracking/ptu_flat_cost_rollup.py
+++ b/litellm/proxy/spend_tracking/ptu_flat_cost_rollup.py
@@ -17,6 +17,7 @@ import json
 from collections.abc import Awaitable, Callable, Mapping
 from dataclasses import dataclass
 from datetime import date, datetime, time, timedelta, timezone
+from types import MappingProxyType
 from typing import TYPE_CHECKING, Final
 
 from litellm._logging import verbose_proxy_logger
@@ -110,17 +111,56 @@ def _public_model_name(row: object, model_info: Mapping[str, object]) -> str:
 
 
 def _decode_model_info(raw: object) -> "Mapping[str, object] | None":
-    """A deployment's model_info as a dict, decoding a JSON string, else None."""
+    """A deployment's model_info as a mapping, decoding a JSON string, else None.
+
+    Valid JSON that is not an object decodes to a list or a scalar, which every caller
+    would then read fields off, so it is rejected here rather than raised past them.
+    """
     if isinstance(raw, str):
         try:
-            return json.loads(raw)
+            decoded: Final = json.loads(raw)
         except (TypeError, ValueError):
             return None
-    if isinstance(raw, dict):
+        return decoded if isinstance(decoded, dict) else None
+    if isinstance(raw, Mapping):
         return raw
     return None
 
 
+@dataclass(frozen=True, slots=True)
+class _PTUDeployment:
+    """A deployment in the shape ``_parse_ptu_model`` reads, whatever declared it.
+
+    A ``LiteLLM_ProxyModelTable`` row already has it. A router entry does not: its id
+    lives in ``model_info.id`` rather than on the entry itself.
+    """
+
+    model_id: str
+    model_name: str
+    model_info: Mapping[str, object]
+
+
+def _router_deployment(deployment: Mapping[str, object]) -> _PTUDeployment | None:
+    """A router ``model_list`` entry in the shape the parser reads, else None.
+
+    An id is required rather than defaulted because it keys the sentinel row: every
+    deployment without one would collapse onto a single row per team and only the last
+    would be billed. The mapping is copied because the router rewrites entries in place
+    while the rollup runs.
+    """
+    model_info: Final = _decode_model_info(deployment.get("model_info"))
+    if model_info is None:
+        return None
+    model_id: Final = model_info.get("id")
+    if not isinstance(model_id, str) or not model_id:
+        return None
+    return _PTUDeployment(
+        model_id=model_id,
+        model_name=str(deployment.get("model_name") or ""),
+        model_info=MappingProxyType(dict(model_info)),
+    )
+
+
 def _parse_ptu_model(row: object) -> PTUModel | None:
     """Return a PTUModel when the deployment carries valid manual PTU config, else None.
 
diff --git a/tests/test_litellm/proxy/spend_tracking/test_ptu_flat_cost_rollup.py b/tests/test_litellm/proxy/spend_tracking/test_ptu_flat_cost_rollup.py
index 736fc13d137..f41756d2b87 100644
--- a/tests/test_litellm/proxy/spend_tracking/test_ptu_flat_cost_rollup.py
+++ b/tests/test_litellm/proxy/spend_tracking/test_ptu_flat_cost_rollup.py
@@ -1,5 +1,6 @@
 """Tests for the per-model PTU flat-cost daily rollup."""
 
+import json
 import types
 from datetime import date, datetime, timedelta, timezone
 from unittest.mock import AsyncMock, MagicMock
@@ -373,6 +374,182 @@ def test_parse_ptu_model_rejects_an_inverted_window():
     assert parsed is None
 
 
+def _router_entry(model_id="dep-1", model_name="gpt-4o-ptu", model_info=None, with_start=True):
+    """A deployment as the router stores one: a plain dict whose id lives in model_info."""
+    info = dict(model_info or {})
+    if (
+        with_start
+        and info.get("ptu_count") is not None
+        and info.get("cost_per_ptu_per_hour") is not None
+        and "ptu_effective_from" not in info
+    ):
+        info["ptu_effective_from"] = _DEFAULT_PTU_START
+    if model_id is not None:
+        info["id"] = model_id
+    return {
+        "model_name": model_name,
+        "litellm_params": {"model": "azure/gpt-4o"},
+        "model_info": info,
+    }
+
+
+# A team-scoped deployment is stored under a synthetic routing key with the operator's name
+# in team_public_model_name, while the same deployment in config.yaml carries the operator's
+# name directly. Both must resolve to the same PTUModel or the two sources bill differently.
+_PARITY_CASES = (
+    (
+        "an iso string start against the datetime pydantic coerces it to",
+        {"ptu_effective_from": "2026-07-30T23:00:00Z"},
+        {"ptu_effective_from": datetime(2026, 7, 30, 23, 0)},
+        datetime(2026, 7, 30, 23, 0, tzinfo=timezone.utc),
+        None,
+    ),
+    (
+        "an open window, stored as null and dropped by exclude_none",
+        {"ptu_effective_from": "2026-07-01T00:00:00Z", "ptu_effective_to": None},
+        {"ptu_effective_from": datetime(2026, 7, 1, 0, 0)},
+        datetime(2026, 7, 1, tzinfo=timezone.utc),
+        None,
+    ),
+    (
+        "a closed window",
+        {"ptu_effective_from": "2026-07-01T00:00:00Z", "ptu_effective_to": "2026-07-31T00:00:00Z"},
+        {
+            "ptu_effective_from": datetime(2026, 7, 1, 0, 0),
+            "ptu_effective_to": datetime(2026, 7, 31, 0, 0),
+        },
+        datetime(2026, 7, 1, tzinfo=timezone.utc),
+        datetime(2026, 7, 31, tzinfo=timezone.utc),
+    ),
+)
+
+
+@pytest.mark.parametrize(
+    "db_window, router_window, expected_from, expected_to",
+    [case[1:] for case in _PARITY_CASES],
+    ids=[case[0] for case in _PARITY_CASES],
+)
+def test_a_deployment_parses_identically_from_the_db_and_from_config_yaml(
+    db_window, router_window, expected_from, expected_to
+):
+    # The contract the router union is built on. If these ever diverge, a PTU deployment
+    # declared in config.yaml is billed differently from the identical one in the database.
+    from_db = _parse_ptu_model(
+        _model_row(
+            model_id="dep-1",
+            model_name="model_name_t_9f3c2b",
+            model_info={**_VALID_PTU, **db_window, "team_public_model_name": "gpt-4o-ptu"},
+            with_start=False,
+        )
+    )
+    from_router = _parse_ptu_model(
+        ptu_rollup._router_deployment(
+            _router_entry(model_id="dep-1", model_info={**_VALID_PTU, **router_window}, with_start=False)
+        )
+    )
+    expected = PTUModel(
+        model_id="dep-1",
+        model_name="gpt-4o-ptu",
+        team_id="t",
+        ptu_count=5,
+        cost_per_ptu_per_hour=2.0,
+        effective_from=expected_from,
+        effective_to=expected_to,
+    )
+    assert from_db == from_router == expected
+
+
+@pytest.mark.parametrize("model_id", [None, "", 12345], ids=["absent", "blank", "not a string"])
+def test_a_router_deployment_without_a_usable_id_is_dropped(model_id):
+    # model_id keys the sentinel row, so an unusable one would file every such deployment
+    # in a team onto one row and bill for a single reservation.
+    entry = _router_entry(model_id=None, model_info=dict(_VALID_PTU))
+    if model_id is not None:
+        entry["model_info"]["id"] = model_id
+    assert ptu_rollup._router_deployment(entry) is None
+
+
+def test_a_router_deployment_keeps_the_name_the_operator_wrote():
+    priced = _parse_ptu_model(
+        ptu_rollup._router_deployment(_router_entry(model_name="gpt-4o-ptu", model_info=dict(_VALID_PTU)))
+    )
+    assert priced is not None
+    assert priced.model_name == "gpt-4o-ptu"
+
+
+def test_a_team_alias_on_a_router_deployment_still_wins_over_the_routing_name():
+    # config.yaml can carry team_public_model_name to expose a team-facing alias, and the
+    # charge has to file under the name that team calls rather than the routing one.
+    priced = _parse_ptu_model(
+        ptu_rollup._router_deployment(
+            _router_entry(
+                model_name="routing-name",
+                model_info={**_VALID_PTU, "team_public_model_name": "public-alias"},
+            )
+        )
+    )
+    assert priced is not None
+    assert priced.model_name == "public-alias"
+
+
+def test_a_router_deployment_with_a_stringified_model_info_still_decodes():
+    entry = _router_entry(model_info=dict(_VALID_PTU))
+    entry["model_info"] = json.dumps(entry["model_info"])
+    parsed = _parse_ptu_model(ptu_rollup._router_deployment(entry))
+    assert parsed is not None
+    assert parsed.model_id == "dep-1"
+
+
+@pytest.mark.parametrize(
+    "model_info",
+    [None, "not json", 42, "[1, 2, 3]", '"a string"', "42", "true", "null"],
+    ids=[
+        "absent",
+        "unparseable",
+        "not a string or dict",
+        "json array",
+        "json string",
+        "json number",
+        "json bool",
+        "json null",
+    ],
+)
+def test_a_router_deployment_without_usable_model_info_is_dropped(model_info):
+    # Valid JSON that is not an object decodes to a list or a scalar, and reading fields
+    # off one raises rather than dropping the single bad deployment the rollup expects
+    assert ptu_rollup._router_deployment({"model_name": "x", "model_info": model_info}) is None
+
+
+@pytest.mark.parametrize(
+    "model_info",
+    ["[1, 2, 3]", '"a string"', "42", "true", "null"],
+    ids=["json array", "json string", "json number", "json bool", "json null"],
+)
+def test_a_db_row_holding_non_object_json_is_dropped(model_info):
+    assert _parse_ptu_model(_model_row(model_info=model_info, with_start=False)) is None
+
+
+def test_a_router_deployment_does_not_alias_the_routers_own_model_info():
+    # The rollup runs on a cron while requests are in flight, and the router rewrites
+    # model_info in place, so a held record must neither observe nor cause those writes.
+    live = {"id": "dep-1", **_VALID_PTU}
+    record = ptu_rollup._router_deployment({"model_name": "x", "model_info": live})
+    assert record is not None
+
+    live["ptu_count"] = 999
+    assert record.model_info["ptu_count"] == _VALID_PTU["ptu_count"]
+
+    with pytest.raises(TypeError):
+        record.model_info["ptu_count"] = 1
+
+
+def test_a_raw_router_dict_is_not_a_deployment_record():
+    # The parser reads attributes, so a router dict passed straight to it returns None
+    # instead of raising. Skipping the factory would silently drop every config.yaml
+    # deployment while leaving the suite green.
+    assert _parse_ptu_model(_router_entry(model_info=dict(_VALID_PTU))) is None
+
+
 @pytest.mark.asyncio
 async def test_rollup_returns_empty_when_prisma_client_is_none():
     result = await run_ptu_flat_cost_rollup(None, target_date=DAY)

From a613773fca0161df9ab879f12f02caaa7464af99 Mon Sep 17 00:00:00 2001
From: tin-berri 
Date: Wed, 19 Aug 2026 14:02:15 -0700
Subject: [PATCH 8/9] feat(auto-router)!: scope shadow eval jobs to multiple
 keys (#37251)

* feat(auto-router): scope shadow eval jobs to multiple keys

A shadow eval job now covers a set of keys instead of exactly one, and each
key carries its own max_turns budget, so one key exhausting its budget leaves
its siblings sampling. The existing job row already is the per-key unit
(api_key_id, max_turns, stopped_at, and the one-active-per-key-and-direction
partial unique index all live on it), so multi-key is grouping rather than
schema surgery: a new group_id column ties N sibling rows written atomically
by one create_many, the API's job id becomes the group id, and pre-existing
jobs backfill group_id = id so their ids keep resolving. The sampler hot path
is untouched; its test file has a zero-line diff

Results come back pooled plus a per-key breakdown and responses list every key
with its own budget, stop state and read-time labels. The dashboard is adapted
minimally to the new shapes (the picker stays single-key and submits a one-key
list); the multi-select picker and per-key table land in the stacked UI PR

* fix(shadow_eval): derive completed from spent budgets and record operator stops

* fix(shadow_eval): stamp stops atomically and freeze counts at the stamp

The stop endpoint wrote stopped_by and stopped_at as two separate updates, so
a failure between them left a job reading stopped while its unstamped legs
kept sampling, and the retry got 400 already stopped. One UPDATE now stamps
stopped_by and every missing stopped_at together, preserving the stopped_at a
leg earned from its own budget via COALESCE

Attempt counts now exclude attempts that land after a leg's stopped_at, so an
in-flight attempt finishing just after an operator stop can never push a
legacy pre-stopped_by job over its budget and flip it from stopped to
completed at read time

* fix(shadow_eval): backfill stopped_by so legacy stops never read as completions

* chore(ui): regenerate api types for the shadow eval stop fields

* fix(shadow_eval): let the stop statement pick one winner under racing stops

Two operators can both pass the derived-status guard in the race window. The
stop UPDATE now claims only legs with stopped_by still null and the endpoint
judges by its row count, so exactly one caller ever gets the 200 and the loser
gets the same already-stopped 400 a late caller gets

* refactor(shadow_eval): make the stop statement the whole state machine

The status guard ran before the UPDATE, so a stop racing the last budgeted
attempt still claimed the job and it read stopped forever instead of
completed. The statement now claims the job only while a leg still samples
inside the window with no stop recorded, and the endpoint reads once after
writing: a racing operator, a same-instant budget spend, and a repeat stop all
get the 400 naming the status the job actually holds. The pre-write guard and
the hand-built response go away

* chore(ui): regenerate api types for the stop route description
---
 .../migration.sql                             |   7 +
 .../migration.sql                             |   5 +
 .../litellm_proxy_extras/schema.prisma        |  30 +-
 .../auto_router_endpoints.py                  | 378 ++++++---
 litellm/proxy/schema.prisma                   |  30 +-
 .../auto_router_endpoints.py                  | 103 ++-
 schema.prisma                                 |  30 +-
 .../test_auto_router_endpoints.py             | 717 ++++++++++++++----
 .../_components/ShadowEvalSection.test.tsx    |  43 +-
 .../_components/ShadowEvalSection.tsx         |  26 +-
 .../_components/useShadowEval.ts              |   1 +
 ui/litellm-dashboard/src/lib/http/schema.d.ts | 118 ++-
 12 files changed, 1117 insertions(+), 371 deletions(-)
 create mode 100644 litellm-proxy-extras/litellm_proxy_extras/migrations/20260817000000_shadow_eval_multi_key/migration.sql
 create mode 100644 litellm-proxy-extras/litellm_proxy_extras/migrations/20260818224500_add_shadow_eval_stopped_by/migration.sql

diff --git a/litellm-proxy-extras/litellm_proxy_extras/migrations/20260817000000_shadow_eval_multi_key/migration.sql b/litellm-proxy-extras/litellm_proxy_extras/migrations/20260817000000_shadow_eval_multi_key/migration.sql
new file mode 100644
index 00000000000..18ef5c40662
--- /dev/null
+++ b/litellm-proxy-extras/litellm_proxy_extras/migrations/20260817000000_shadow_eval_multi_key/migration.sql
@@ -0,0 +1,7 @@
+ALTER TABLE "LiteLLM_ShadowEvalJob" ADD COLUMN IF NOT EXISTS "group_id" TEXT;
+
+UPDATE "LiteLLM_ShadowEvalJob" SET "group_id" = "id" WHERE "group_id" IS NULL;
+
+ALTER TABLE "LiteLLM_ShadowEvalJob" ALTER COLUMN "group_id" SET NOT NULL;
+
+CREATE INDEX IF NOT EXISTS "LiteLLM_ShadowEvalJob_group_id_idx" ON "LiteLLM_ShadowEvalJob"("group_id");
diff --git a/litellm-proxy-extras/litellm_proxy_extras/migrations/20260818224500_add_shadow_eval_stopped_by/migration.sql b/litellm-proxy-extras/litellm_proxy_extras/migrations/20260818224500_add_shadow_eval_stopped_by/migration.sql
new file mode 100644
index 00000000000..9efa3fdd052
--- /dev/null
+++ b/litellm-proxy-extras/litellm_proxy_extras/migrations/20260818224500_add_shadow_eval_stopped_by/migration.sql
@@ -0,0 +1,5 @@
+-- AlterTable
+ALTER TABLE "LiteLLM_ShadowEvalJob" ADD COLUMN     "stopped_by" TEXT;
+
+UPDATE "LiteLLM_ShadowEvalJob" SET stopped_by = 'unknown'
+WHERE stopped_at IS NOT NULL AND ends_at > (NOW() AT TIME ZONE 'utc');
diff --git a/litellm-proxy-extras/litellm_proxy_extras/schema.prisma b/litellm-proxy-extras/litellm_proxy_extras/schema.prisma
index 52fb447157b..f79e2bb0c18 100644
--- a/litellm-proxy-extras/litellm_proxy_extras/schema.prisma
+++ b/litellm-proxy-extras/litellm_proxy_extras/schema.prisma
@@ -1467,28 +1467,38 @@ model LiteLLM_AutoRouterSession {
   @@index([last_turn_at], map: "idx_autorouter_session_last_turn")
 }
 
-// Shadow eval: evaluation of an auto-router against a key's live traffic, in either
-// direction. forward duplicates the requests the key did not route through the router
-// through it, answering whether the key should adopt it; reverse duplicates the requests
-// the router did serve against a fixed baseline model, answering whether a key already on
-// it still benefits. Either way a sampled slice runs in a detached task and an LLM judge
-// compares real vs shadow responses blind. The job row is immutable config plus
-// stopped_at; every count, status, and spend figure is derived from the append-only
-// attempt rows, so nothing can disagree across pods or stop races.
+// Shadow eval: evaluation of an auto-router against one or more keys' live traffic, in
+// either direction. forward duplicates the requests the keys did not route through the
+// router through it, answering whether they should adopt it; reverse duplicates the
+// requests the router did serve against a fixed baseline model, answering whether a key
+// already on it still benefits. Either way a sampled slice runs in a detached task and an
+// LLM judge compares real vs shadow responses blind. Each row is ONE key's leg of a job:
+// immutable config plus that key's own turn budget and stop state, so one key exhausting
+// its budget never ends a sibling's sampling. A job is the set of legs sharing group_id
+// (the id the API reports), written together by one atomic create_many with identical
+// config; single-key jobs predating group_id were backfilled group_id = id. "One active
+// job per (key, direction)" is a partial unique index on (api_key_id, direction) WHERE
+// stopped_at IS NULL, expressed only in the migration because schema.prisma cannot state
+// partial indexes; it is what makes a concurrent start on another pod race-safe rather
+// than read-then-create. Every count, status, and spend figure is derived from the
+// append-only attempt rows, so nothing can disagree across pods or stop races.
 model LiteLLM_ShadowEvalJob {
   id                String   @id @default(cuid())
-  api_key_id        String   // hashed virtual key whose traffic is shadowed
+  group_id          String   // legs of one job share this; the API's job id
+  api_key_id        String   // hashed virtual key whose traffic this leg shadows
   router_name       String   // the auto-router under evaluation, in either direction
   direction         String   @default("forward") // forward | reverse
   baseline_model    String?  // reverse only: the fixed model the router is judged against
   judge_model       String
   shadow_percentage Float
-  max_turns         Int      // sample budget: judge at most this many turns
+  max_turns         Int      // this key's sample budget: judge at most this many turns
   created_at        DateTime @default(now())
   created_by        String?
   ends_at           DateTime
   stopped_at        DateTime?
+  stopped_by        String?  // operator who stopped it early; null when it ended on its own
 
+  @@index([group_id])
   @@index([api_key_id])
   @@index([created_at])
 }
diff --git a/litellm/proxy/management_endpoints/auto_router_endpoints.py b/litellm/proxy/management_endpoints/auto_router_endpoints.py
index d8b7414f32c..0112ad1f6ed 100644
--- a/litellm/proxy/management_endpoints/auto_router_endpoints.py
+++ b/litellm/proxy/management_endpoints/auto_router_endpoints.py
@@ -6,10 +6,13 @@ POST /auto_router/test_routing - Route one prompt through an unsaved complexity-
 
 from collections.abc import Mapping, Sequence
 from datetime import datetime, timedelta, timezone
+from itertools import groupby
+from operator import attrgetter
 from types import MappingProxyType
 from typing import TYPE_CHECKING, Annotated, Final, Protocol
+from uuid import uuid4
 
-from pydantic import BaseModel, TypeAdapter
+from pydantic import BaseModel, ConfigDict, TypeAdapter, field_validator
 
 from litellm._logging import verbose_proxy_logger
 from litellm.exceptions import BudgetExceededError
@@ -41,6 +44,8 @@ from litellm.types.management_endpoints.auto_router_endpoints import (
     AutoRouterRoutingTestRequest,
     AutoRouterRoutingTestResponse,
     RequestComplexityRouterConfig,
+    ShadowEvalDirection,
+    ShadowEvalJobKeyResponse,
     ShadowEvalJobResponse,
     ShadowEvalResult,
     ShadowEvalSlice,
@@ -89,17 +94,9 @@ class _ShadowEvalJobRow(Protocol):
 
 
 class _ShadowEvalJobTable(Protocol):
-    async def find_unique(self, *, where: Mapping[str, object]) -> _ShadowEvalJobRow | None: ...
+    async def find_many(self, *, where: Mapping[str, object]) -> Sequence[_ShadowEvalJobRow]: ...
 
-    async def find_first(self, *, where: Mapping[str, object]) -> _ShadowEvalJobRow | None: ...
-
-    async def find_many(
-        self, *, where: Mapping[str, object], order: Mapping[str, str], take: int
-    ) -> Sequence[_ShadowEvalJobRow]: ...
-
-    async def create(self, data: Mapping[str, object]) -> _ShadowEvalJobRow: ...
-
-    async def update(self, *, where: Mapping[str, object], data: Mapping[str, object]) -> _ShadowEvalJobRow | None: ...
+    async def create_many(self, data: Sequence[Mapping[str, object]]) -> int: ...
 
 
 class _ShadowEvalAttemptRow(Protocol):
@@ -606,18 +603,19 @@ _ATTEMPT_AGG_SELECT: Final = """
     COUNT(*) FILTER (WHERE outcome = 'tie')::int AS ties,
     AVG(confidence)::float AS avg_confidence
 FROM "LiteLLM_ShadowEvalAttempt"
-WHERE job_id = $1 AND outcome != 'error'
+WHERE job_id = ANY($1::text[]) AND outcome != 'error'
 GROUP BY 1
 """
 
 _ATTEMPT_AGG_BY_TIER_SQL: Final = "SELECT COALESCE(tier, 'UNCLASSIFIED') AS grp," + _ATTEMPT_AGG_SELECT
 _ATTEMPT_AGG_BY_MODEL_SQL: Final = "SELECT COALESCE(real_model, 'unknown') AS grp," + _ATTEMPT_AGG_SELECT
+_ATTEMPT_AGG_BY_LEG_SQL: Final = "SELECT job_id AS grp," + _ATTEMPT_AGG_SELECT
 
 _SWEEP_FINISHED_JOBS_SQL: Final = """
-UPDATE "LiteLLM_ShadowEvalJob" j SET stopped_at = NOW()
-WHERE j.api_key_id = $1 AND j.stopped_at IS NULL
+UPDATE "LiteLLM_ShadowEvalJob" j SET stopped_at = (NOW() AT TIME ZONE 'utc')
+WHERE j.api_key_id = ANY($1::text[]) AND j.stopped_at IS NULL
   AND (
-    j.ends_at <= NOW()
+    j.ends_at <= (NOW() AT TIME ZONE 'utc')
     OR (SELECT COUNT(*) FROM "LiteLLM_ShadowEvalAttempt" a WHERE a.job_id = j.id) >= j.max_turns
   )
 """
@@ -628,7 +626,52 @@ SELECT
     COUNT(*) FILTER (WHERE outcome = 'error')::int AS error_count,
     COALESCE(SUM(judge_cost), 0)::float AS judge_spend
 FROM "LiteLLM_ShadowEvalAttempt"
-WHERE job_id = $1
+WHERE job_id = ANY($1::text[])
+"""
+
+_ATTEMPT_COUNTS_SQL: Final = """
+SELECT a.job_id, COUNT(*)::int AS attempt_count
+FROM "LiteLLM_ShadowEvalAttempt" a
+JOIN "LiteLLM_ShadowEvalJob" j ON j.id = a.job_id
+WHERE a.job_id = ANY($1::text[]) AND (j.stopped_at IS NULL OR a.created_at <= j.stopped_at)
+GROUP BY a.job_id
+"""
+
+_STOP_JOB_SQL: Final = """
+UPDATE "LiteLLM_ShadowEvalJob"
+SET stopped_by = $2, stopped_at = COALESCE(stopped_at, $3::timestamp)
+WHERE group_id = $1 AND stopped_by IS NULL
+  AND ends_at > (NOW() AT TIME ZONE 'utc')
+  AND EXISTS (
+    SELECT 1 FROM "LiteLLM_ShadowEvalJob" k
+    WHERE k.group_id = $1 AND k.stopped_at IS NULL
+      AND (SELECT COUNT(*) FROM "LiteLLM_ShadowEvalAttempt" a WHERE a.job_id = k.id) < k.max_turns
+  )
+"""
+
+
+class _AttemptCountRow(BaseModel):
+    job_id: str
+    attempt_count: int
+
+
+_ATTEMPT_COUNT_ROWS: Final = TypeAdapter(list[_AttemptCountRow])
+
+
+_LIST_LEGS_SQL: Final = """
+SELECT * FROM "LiteLLM_ShadowEvalJob"
+WHERE group_id IN (
+    SELECT group_id FROM "LiteLLM_ShadowEvalJob"
+    GROUP BY group_id ORDER BY MAX(created_at) DESC LIMIT $1::int
+)
+"""
+
+_LIST_LEGS_BY_KEY_SQL: Final = """
+SELECT * FROM "LiteLLM_ShadowEvalJob"
+WHERE group_id IN (
+    SELECT group_id FROM "LiteLLM_ShadowEvalJob" WHERE api_key_id = $2
+    GROUP BY group_id ORDER BY MAX(created_at) DESC LIMIT $1::int
+)
 """
 
 
@@ -659,18 +702,98 @@ def _slices(rows: Sequence[_AttemptAggRow]) -> tuple[ShadowEvalSlice, ...]:
     )
 
 
+class _LegRow(BaseModel):
+    """One LiteLLM_ShadowEvalJob row, validated off the untyped prisma record. A row is
+    one key's leg of a job; the legs of a job share group_id and identical config, written
+    together by one create_many. The API's job id is the group id, so leg ids never leave
+    the server (attempts reference them internally)."""
+
+    model_config = ConfigDict(from_attributes=True)
+
+    id: str
+    group_id: str
+    api_key_id: str
+    router_name: str
+    direction: ShadowEvalDirection
+    baseline_model: str | None = None
+    judge_model: str
+    shadow_percentage: float
+    max_turns: int
+    created_at: datetime
+    ends_at: datetime
+    stopped_at: datetime | None = None
+    stopped_by: str | None = None
+
+    @field_validator("created_at", "ends_at", "stopped_at")
+    @classmethod
+    def _as_aware_utc(cls, value: datetime | None) -> datetime | None:
+        """The columns store naive UTC wall time (prisma's convention); prisma reads hand
+        back aware datetimes while raw SQL reads hand back naive ones, so this boundary
+        makes every read aware UTC before anything compares or serializes them."""
+        if value is None or value.tzinfo is not None:
+            return value
+        return value.replace(tzinfo=timezone.utc)
+
+
+_LEG_ROWS: Final = TypeAdapter(list[_LegRow])
+
+
+async def _leg_attempt_counts(prisma_client: "PrismaClient", legs: Sequence[_LegRow]) -> Mapping[str, int]:
+    """Each leg's attempt count by leg id, judged and errored alike, in one grouped read.
+    It is the same count the sampler budgets against max_turns, so the derived status
+    flips to completed exactly when sampling actually ends. A stamped leg's count freezes
+    at its stopped_at: in-flight attempts that land after the stamp are excluded, so they
+    can never reclassify a leg that was stopped under budget as budget-spent."""
+    if not legs:
+        return MappingProxyType({})
+    rows: Final = _ATTEMPT_COUNT_ROWS.validate_python(
+        await _query_raw(prisma_client, _ATTEMPT_COUNTS_SQL, [leg.id for leg in legs])  # mutable-ok: query param
+        or ()
+    )
+    return MappingProxyType({row.job_id: row.attempt_count for row in rows})
+
+
+def _group_response(group_id: str, legs: Sequence[_LegRow], attempt_counts: Mapping[str, int]) -> ShadowEvalJobResponse:
+    """The one constructor of a job response: the caller names the group and passes that
+    group's legs. Config is read off the first leg because every leg carries the same copy,
+    written by one create_many. No caller may serialize a raw row (that would leak a leg id
+    as the job id)."""
+    first: Final = legs[0]
+    return ShadowEvalJobResponse(
+        job_id=group_id,
+        keys=tuple(
+            ShadowEvalJobKeyResponse(
+                api_key_id=leg.api_key_id,
+                max_turns=leg.max_turns,
+                stopped_at=leg.stopped_at,
+                attempt_count=attempt_counts.get(leg.id, 0),
+            )
+            for leg in sorted(legs, key=lambda leg: leg.api_key_id)
+        ),
+        router_name=first.router_name,
+        direction=first.direction,
+        baseline_model=first.baseline_model,
+        judge_model=first.judge_model,
+        shadow_percentage=first.shadow_percentage,
+        created_at=first.created_at,
+        ends_at=first.ends_at,
+        stopped_by=next((leg.stopped_by for leg in legs if leg.stopped_by is not None), None),
+    )
+
+
 _NO_KEY_LABELS: Final[tuple[str | None, str | None]] = (None, None)
 
 
 async def _with_key_labels(
     prisma_client: "PrismaClient", responses: Sequence[ShadowEvalJobResponse]
 ) -> tuple[ShadowEvalJobResponse, ...]:
-    """Resolve each job's key hash to the key's alias and masked name in one batched read,
+    """Resolve every scoped key's hash to its alias and masked name in one batched read,
     so the UI can say whose traffic a job shadows. Deleted keys resolve to None."""
     if not responses:
         return ()
+    tokens: Final = sorted(frozenset(key.api_key_id for response in responses for key in response.keys))
     key_rows: Final = await _verification_tokens(prisma_client).find_many(
-        where={"token": {"in": sorted({response.api_key_id for response in responses})}}  # mutable-ok: Prisma filter
+        where={"token": {"in": tokens}}  # mutable-ok: Prisma filter
     )
     labels: Final[Mapping[str, tuple[str | None, str | None]]] = {
         row.token: (row.key_alias, row.key_name) for row in key_rows or ()
@@ -678,32 +801,50 @@ async def _with_key_labels(
     return tuple(
         response.model_copy(
             update={  # mutable-ok: pydantic update payload
-                "key_alias": labels.get(response.api_key_id, _NO_KEY_LABELS)[0],
-                "key_name": labels.get(response.api_key_id, _NO_KEY_LABELS)[1],
+                "keys": tuple(
+                    key.model_copy(
+                        update={  # mutable-ok: pydantic update payload
+                            "key_alias": labels.get(key.api_key_id, _NO_KEY_LABELS)[0],
+                            "key_name": labels.get(key.api_key_id, _NO_KEY_LABELS)[1],
+                        }
+                    )
+                    for key in response.keys
+                )
             }
         )
         for response in responses
     )
 
 
-async def _shadow_eval_results(prisma_client: "PrismaClient", job_id: str) -> ShadowEvalResult | None:
-    """Both stratifications of one job's verdicts. Tier answers "where does the router do
-    well"; the model stratification groups by whichever model served the real arm, so it
-    answers "which of the models this key uses today would the router beat" forward, and
-    "for the turns the router sent to X, did X beat the baseline" in reverse. Reads are
-    bounded by the job's own attempts (<= max_turns) via the job_id index."""
+async def _shadow_eval_results(prisma_client: "PrismaClient", legs: Sequence[_LegRow]) -> ShadowEvalResult | None:
+    """All three stratifications of one job's verdicts. Tier answers "where does the router
+    do well"; the model stratification groups by whichever model served the real arm, so it
+    answers "which of the models these keys use today would the router beat" forward, and
+    "for the turns the router sent to X, did X beat the baseline" in reverse; key answers
+    "which key's traffic does the router suit". Reads are bounded by the job's own attempts
+    (<= the sum of its keys' max_turns) via the job_id index."""
+    leg_ids: Final = [leg.id for leg in legs]  # mutable-ok: query param
     by_tier: Final = _ATTEMPT_AGG_ROWS.validate_python(
-        await _query_raw(prisma_client, _ATTEMPT_AGG_BY_TIER_SQL, job_id) or ()
+        await _query_raw(prisma_client, _ATTEMPT_AGG_BY_TIER_SQL, leg_ids) or ()
     )
     if not by_tier:
         return None
     by_model: Final = _ATTEMPT_AGG_ROWS.validate_python(
-        await _query_raw(prisma_client, _ATTEMPT_AGG_BY_MODEL_SQL, job_id) or ()
+        await _query_raw(prisma_client, _ATTEMPT_AGG_BY_MODEL_SQL, leg_ids) or ()
+    )
+    key_by_leg: Final = MappingProxyType({leg.id: leg.api_key_id for leg in legs})
+    by_leg: Final = _ATTEMPT_AGG_ROWS.validate_python(
+        await _query_raw(prisma_client, _ATTEMPT_AGG_BY_LEG_SQL, leg_ids) or ()
+    )
+    by_key: Final = tuple(
+        row.model_copy(update={"grp": key_by_leg[row.grp]})  # mutable-ok: pydantic update payload
+        for row in by_leg
     )
     total_turns: Final = sum(r.turn_count for r in by_tier)
     return ShadowEvalResult(
         by_tier=_slices(by_tier),
         by_current_model=_slices(by_model),
+        by_key=_slices(by_key),
         overall_shadow_win_rate_pct=_pct_of(sum(r.shadow_wins for r in by_tier), total_turns),
         overall_tie_rate_pct=_pct_of(sum(r.ties for r in by_tier), total_turns),
     )
@@ -721,20 +862,21 @@ async def start_shadow_eval(
     user_api_key_dict: Annotated[UserAPIKeyAuth, Depends(user_api_key_auth)],
 ) -> ShadowEvalJobResponse:
     """
-    Start a shadow eval: duplicate a sampled slice of a key's live traffic against a second
-    arm, judge the two responses blind, and stratify win rates by tier and by the model that
-    served the real arm.
+    Start a shadow eval: duplicate a sampled slice of one or more keys' live traffic against
+    a second arm, judge the two responses blind, and stratify win rates by tier, by the model
+    that served the real arm, and by key.
 
-    A forward job answers whether the key should adopt router_name: it samples the requests
+    A forward job answers whether the keys should adopt router_name: it samples the requests
     the router did not serve and duplicates them through it. A reverse job answers whether a
     key already on the router still gains from it: it samples the requests the router did
     serve and duplicates them against baseline_model. A key can hold one active job per
     direction, so both questions can run at once.
 
-    Shadow responses are never served to users. The job samples until it has judged
-    max_turns turns, reaches the end of its window, or is stopped; sampling changes
-    propagate to pods within about 10 seconds. Shadow and judge calls bill to the
-    shadowed key but are excluded from request counts and auto-router adoption metrics.
+    Shadow responses are never served to users. Each key samples until it has judged
+    max_turns turns of its own traffic, the job's window ends, or the job is stopped, so one
+    key running out of budget does not end sampling for the others; sampling changes
+    propagate to pods within about 10 seconds. Shadow and judge calls bill to the shadowed
+    key but are excluded from request counts and auto-router adoption metrics.
     """
     from litellm.proxy.proxy_server import llm_router, prisma_client
 
@@ -746,48 +888,58 @@ async def start_shadow_eval(
     _validate_plain_model(llm_router, data.judge_model, "judge_model")
     if data.baseline_model is not None:
         _validate_plain_model(llm_router, data.baseline_model, "baseline_model")
-    key_row: Final = await _verification_tokens(prisma_client).find_unique(
-        where={"token": data.api_key_id}  # mutable-ok: Prisma filter
+    token_rows: Final = await _verification_tokens(prisma_client).find_many(
+        where={"token": {"in": list(data.api_key_ids)}}  # mutable-ok: Prisma filter
     )
-    if key_row is None:
+    unknown: Final = tuple(sorted(frozenset(data.api_key_ids) - frozenset(row.token for row in token_rows or ())))
+    if unknown:
         raise HTTPException(
             status_code=400,
             detail=(
-                f"api_key_id '{data.api_key_id}' is not a key on this proxy; pass the key's token hash, "
+                f"api_key_ids not on this proxy: {', '.join(unknown)}; pass each key's token hash, "
                 "the value the key list and key info endpoints report"
             ),
         )
 
-    # A job that expired or exhausted its turn budget stopped sampling on its own, but
-    # still holds its slot in the per-key, per-direction partial unique index until
-    # stamped; free it so a new eval can start. Sweeping both directions is deliberate.
-    await prisma_client.db.execute_raw(_SWEEP_FINISHED_JOBS_SQL, data.api_key_id)
-    active: Final = await _shadow_eval_jobs(prisma_client).find_first(
+    # A job whose window passed or whose turn budget ran out stopped sampling on its own,
+    # but its legs still hold their slots in the per-key, per-direction partial unique index
+    # until stamped; free them so a new eval can start. Sweeping both directions is deliberate.
+    requested: Final = list(data.api_key_ids)  # mutable-ok: query param
+    await prisma_client.db.execute_raw(_SWEEP_FINISHED_JOBS_SQL, requested)
+    claimed: Final = await _shadow_eval_jobs(prisma_client).find_many(
         where={  # mutable-ok: Prisma filter
-            "api_key_id": data.api_key_id,
+            "api_key_id": {"in": requested},  # mutable-ok: Prisma filter
             "direction": data.direction,
             "stopped_at": None,
         },
     )
-    if active is not None:
+    if claimed:
         raise HTTPException(
             status_code=409,
-            detail=f"Key already has an active {data.direction} shadow eval job ({active.id}). Stop it first.",
+            detail=(
+                f"Already in an active {data.direction} shadow eval job: "
+                + ", ".join(sorted(f"{row.api_key_id} (job {row.group_id})" for row in claimed))
+                + ". Stop it first."
+            ),
         )
     now: Final = datetime.now(timezone.utc)
+    group_id: Final = str(uuid4())
+    ends_at: Final = now + timedelta(days=data.duration_days)
+    shared_config: Final = {  # mutable-ok: Prisma payload
+        "group_id": group_id,
+        "router_name": data.router_name,
+        "direction": data.direction,
+        "baseline_model": data.baseline_model,
+        "judge_model": data.judge_model,
+        "shadow_percentage": data.shadow_percentage,
+        "max_turns": data.max_turns,
+        "created_by": user_api_key_dict.user_id,
+        "created_at": now,
+        "ends_at": ends_at,
+    }
     try:
-        job: Final = await _shadow_eval_jobs(prisma_client).create(
-            data={  # mutable-ok: Prisma payload
-                "api_key_id": data.api_key_id,
-                "router_name": data.router_name,
-                "direction": data.direction,
-                "baseline_model": data.baseline_model,
-                "judge_model": data.judge_model,
-                "shadow_percentage": data.shadow_percentage,
-                "max_turns": data.max_turns,
-                "created_by": user_api_key_dict.user_id,
-                "ends_at": now + timedelta(days=data.duration_days),
-            }
+        await _shadow_eval_jobs(prisma_client).create_many(
+            data=[{**shared_config, "api_key_id": key} for key in data.api_key_ids]  # mutable-ok: Prisma payload
         )
     except Exception as e:
         if not _is_unique_violation(e):
@@ -795,11 +947,28 @@ async def start_shadow_eval(
         raise HTTPException(
             status_code=409,
             detail=(
-                f"Key already has an active {data.direction} shadow eval job (started concurrently). Stop it first."
+                f"A requested key was claimed by another {data.direction} shadow eval job concurrently. Stop it first."
             ),
         ) from e
-    return ShadowEvalJobResponse.model_validate(job, from_attributes=True).model_copy(
-        update={"key_alias": key_row.key_alias, "key_name": key_row.key_name}  # mutable-ok: pydantic update payload
+    labels: Final = MappingProxyType({row.token: row for row in token_rows})
+    return ShadowEvalJobResponse(
+        job_id=group_id,
+        keys=tuple(
+            ShadowEvalJobKeyResponse(
+                api_key_id=api_key_id,
+                max_turns=data.max_turns,
+                key_alias=labels[api_key_id].key_alias,
+                key_name=labels[api_key_id].key_name,
+            )
+            for api_key_id in sorted(data.api_key_ids)
+        ),
+        router_name=data.router_name,
+        direction=data.direction,
+        baseline_model=data.baseline_model,
+        judge_model=data.judge_model,
+        shadow_percentage=data.shadow_percentage,
+        created_at=now,
+        ends_at=ends_at,
     )
 
 
@@ -811,23 +980,38 @@ async def start_shadow_eval(
 )
 async def list_shadow_eval_jobs(
     user_api_key_dict: Annotated[UserAPIKeyAuth, Depends(user_api_key_auth)],
-    api_key_id: Annotated[str | None, Query(description="Filter to jobs shadowing this key")] = None,
+    api_key_id: Annotated[
+        str | None, Query(description="Filter to jobs that shadow this key, alone or alongside others")
+    ] = None,
     limit: Annotated[int, Query(ge=1, le=200, description="Newest jobs to return")] = 50,
 ) -> tuple[ShadowEvalJobResponse, ...]:
-    """List shadow eval jobs, newest first. Counts and results ride the detail endpoint only."""
+    """List shadow eval jobs, newest first, each key with its attempt count so status is
+    accurate. Judged counts, spend, and results ride the detail endpoint only."""
     from litellm.proxy.proxy_server import prisma_client
 
     _require_admin_viewer(user_api_key_dict, "view shadow evals")
     if prisma_client is None:
         raise HTTPException(status_code=500, detail=CommonProxyErrors.db_not_connected_error.value)
-    records: Final = await _shadow_eval_jobs(prisma_client).find_many(
-        where={"api_key_id": api_key_id} if api_key_id else {},  # mutable-ok: Prisma filter
-        order={"created_at": "desc"},  # mutable-ok: Prisma order
-        take=limit,
+    legs: Final = _LEG_ROWS.validate_python(
+        (
+            await _query_raw(prisma_client, _LIST_LEGS_BY_KEY_SQL, limit, api_key_id)
+            if api_key_id
+            else await _query_raw(prisma_client, _LIST_LEGS_SQL, limit)
+        )
+        or ()
     )
+    by_group: Final[Mapping[str, tuple[_LegRow, ...]]] = MappingProxyType(
+        {
+            group_id: tuple(group)
+            for group_id, group in groupby(sorted(legs, key=attrgetter("group_id")), key=attrgetter("group_id"))
+        }
+    )
+    newest_first: Final = sorted(
+        by_group, key=lambda group_id: max(leg.created_at for leg in by_group[group_id]), reverse=True
+    )
+    counts: Final = await _leg_attempt_counts(prisma_client, legs)
     return await _with_key_labels(
-        prisma_client,
-        tuple(ShadowEvalJobResponse.model_validate(record, from_attributes=True) for record in records or ()),
+        prisma_client, tuple(_group_response(group_id, by_group[group_id], counts) for group_id in newest_first)
     )
 
 
@@ -847,20 +1031,24 @@ async def get_shadow_eval_job(
     _require_admin_viewer(user_api_key_dict, "view shadow evals")
     if prisma_client is None:
         raise HTTPException(status_code=500, detail=CommonProxyErrors.db_not_connected_error.value)
-    record: Final = await _shadow_eval_jobs(prisma_client).find_unique(
-        where={"id": job_id}  # mutable-ok: Prisma filter
+    legs: Final = _LEG_ROWS.validate_python(
+        await _shadow_eval_jobs(prisma_client).find_many(
+            where={"group_id": job_id}  # mutable-ok: Prisma filter
+        )
+        or ()
     )
-    if record is None:
+    if not legs:
         raise HTTPException(status_code=404, detail=f"No shadow eval job {job_id}")
+    leg_ids: Final = [leg.id for leg in legs]  # mutable-ok: query param
     totals: Final = _ATTEMPT_TOTALS_ROWS.validate_python(
-        await _query_raw(prisma_client, _ATTEMPT_TOTALS_SQL, job_id) or ()
+        await _query_raw(prisma_client, _ATTEMPT_TOTALS_SQL, leg_ids) or ()
     )
     latest_error: Final = await _shadow_eval_attempts(prisma_client).find_first(
-        where={"job_id": job_id, "outcome": "error"},  # mutable-ok: Prisma filter
+        where={"job_id": {"in": leg_ids}, "outcome": "error"},  # mutable-ok: Prisma filter
         order={"created_at": "desc"},  # mutable-ok: Prisma order
     )
     labeled: Final = await _with_key_labels(
-        prisma_client, (ShadowEvalJobResponse.model_validate(record, from_attributes=True),)
+        prisma_client, (_group_response(job_id, legs, await _leg_attempt_counts(prisma_client, legs)),)
     )
     return labeled[0].model_copy(
         update={  # mutable-ok: pydantic update payload
@@ -868,7 +1056,7 @@ async def get_shadow_eval_job(
             "error_count": totals[0].error_count if totals else 0,
             "judge_spend": round(totals[0].judge_spend, 6) if totals else 0.0,
             "last_error": latest_error.error if latest_error else None,
-            "results": await _shadow_eval_results(prisma_client, job_id),
+            "results": await _shadow_eval_results(prisma_client, legs),
         }
     )
 
@@ -883,25 +1071,33 @@ async def stop_shadow_eval_job(
     job_id: str,
     user_api_key_dict: Annotated[UserAPIKeyAuth, Depends(user_api_key_auth)],
 ) -> ShadowEvalJobResponse:
-    """Stop an active shadow eval job. Attempts are kept; sampling halts within ~10s."""
+    """Stop an active shadow eval job, every key it scopes at once. Attempts are kept;
+    sampling halts within ~10s. Keys that already stopped on their own budget keep the
+    stopped_at they earned. The statement is the whole state machine: it claims the job
+    only while a leg still samples inside the window with no stop recorded, so a racing
+    operator, a same-instant budget spend, and a repeat stop all read the same 400 with
+    the status the job actually holds."""
     from litellm.proxy.proxy_server import prisma_client
 
     _require_admin_writer(user_api_key_dict, "stop a shadow eval")
     if prisma_client is None:
         raise HTTPException(status_code=500, detail=CommonProxyErrors.db_not_connected_error.value)
-    record: Final = await _shadow_eval_jobs(prisma_client).find_unique(
-        where={"id": job_id}  # mutable-ok: Prisma filter
+    stamp: Final = datetime.now(timezone.utc)
+    operator: Final = user_api_key_dict.user_id or "operator"
+    claimed: Final = await prisma_client.db.execute_raw(
+        _STOP_JOB_SQL, job_id, operator, stamp.replace(tzinfo=None).isoformat()
     )
-    if record is None:
+    legs: Final = _LEG_ROWS.validate_python(
+        await _shadow_eval_jobs(prisma_client).find_many(
+            where={"group_id": job_id}  # mutable-ok: Prisma filter
+        )
+        or ()
+    )
+    if not legs:
         raise HTTPException(status_code=404, detail=f"No shadow eval job {job_id}")
-    current: Final = ShadowEvalJobResponse.model_validate(record, from_attributes=True)
-    if current.status != "running":
+    counts: Final = await _leg_attempt_counts(prisma_client, legs)
+    current: Final = _group_response(job_id, legs, counts)
+    if claimed == 0:
         raise HTTPException(status_code=400, detail=f"Job {job_id} is already {current.status}")
-    updated: Final = await _shadow_eval_jobs(prisma_client).update(
-        where={"id": job_id},  # mutable-ok: Prisma filter
-        data={"stopped_at": datetime.now(timezone.utc)},  # mutable-ok: Prisma payload
-    )
-    labeled: Final = await _with_key_labels(
-        prisma_client, (ShadowEvalJobResponse.model_validate(updated, from_attributes=True),)
-    )
+    labeled: Final = await _with_key_labels(prisma_client, (current,))
     return labeled[0]
diff --git a/litellm/proxy/schema.prisma b/litellm/proxy/schema.prisma
index 52fb447157b..f79e2bb0c18 100644
--- a/litellm/proxy/schema.prisma
+++ b/litellm/proxy/schema.prisma
@@ -1467,28 +1467,38 @@ model LiteLLM_AutoRouterSession {
   @@index([last_turn_at], map: "idx_autorouter_session_last_turn")
 }
 
-// Shadow eval: evaluation of an auto-router against a key's live traffic, in either
-// direction. forward duplicates the requests the key did not route through the router
-// through it, answering whether the key should adopt it; reverse duplicates the requests
-// the router did serve against a fixed baseline model, answering whether a key already on
-// it still benefits. Either way a sampled slice runs in a detached task and an LLM judge
-// compares real vs shadow responses blind. The job row is immutable config plus
-// stopped_at; every count, status, and spend figure is derived from the append-only
-// attempt rows, so nothing can disagree across pods or stop races.
+// Shadow eval: evaluation of an auto-router against one or more keys' live traffic, in
+// either direction. forward duplicates the requests the keys did not route through the
+// router through it, answering whether they should adopt it; reverse duplicates the
+// requests the router did serve against a fixed baseline model, answering whether a key
+// already on it still benefits. Either way a sampled slice runs in a detached task and an
+// LLM judge compares real vs shadow responses blind. Each row is ONE key's leg of a job:
+// immutable config plus that key's own turn budget and stop state, so one key exhausting
+// its budget never ends a sibling's sampling. A job is the set of legs sharing group_id
+// (the id the API reports), written together by one atomic create_many with identical
+// config; single-key jobs predating group_id were backfilled group_id = id. "One active
+// job per (key, direction)" is a partial unique index on (api_key_id, direction) WHERE
+// stopped_at IS NULL, expressed only in the migration because schema.prisma cannot state
+// partial indexes; it is what makes a concurrent start on another pod race-safe rather
+// than read-then-create. Every count, status, and spend figure is derived from the
+// append-only attempt rows, so nothing can disagree across pods or stop races.
 model LiteLLM_ShadowEvalJob {
   id                String   @id @default(cuid())
-  api_key_id        String   // hashed virtual key whose traffic is shadowed
+  group_id          String   // legs of one job share this; the API's job id
+  api_key_id        String   // hashed virtual key whose traffic this leg shadows
   router_name       String   // the auto-router under evaluation, in either direction
   direction         String   @default("forward") // forward | reverse
   baseline_model    String?  // reverse only: the fixed model the router is judged against
   judge_model       String
   shadow_percentage Float
-  max_turns         Int      // sample budget: judge at most this many turns
+  max_turns         Int      // this key's sample budget: judge at most this many turns
   created_at        DateTime @default(now())
   created_by        String?
   ends_at           DateTime
   stopped_at        DateTime?
+  stopped_by        String?  // operator who stopped it early; null when it ended on its own
 
+  @@index([group_id])
   @@index([api_key_id])
   @@index([created_at])
 }
diff --git a/litellm/types/management_endpoints/auto_router_endpoints.py b/litellm/types/management_endpoints/auto_router_endpoints.py
index 9461297feca..63c93e0f268 100644
--- a/litellm/types/management_endpoints/auto_router_endpoints.py
+++ b/litellm/types/management_endpoints/auto_router_endpoints.py
@@ -6,7 +6,7 @@ from collections.abc import Mapping
 from datetime import datetime, timezone
 from typing import Final, Literal, TypeAlias
 
-from pydantic import AliasChoices, BaseModel, ConfigDict, Field, computed_field, field_validator, model_validator
+from pydantic import BaseModel, Field, computed_field, field_validator, model_validator
 
 from litellm.router_strategy.complexity_router.config import ComplexityRouterConfig
 from litellm.types.utils import StandardLoggingRoutingDecision
@@ -155,13 +155,17 @@ DEFAULT_SHADOW_EVAL_JUDGE_MODEL: Final[str] = "anthropic/claude-sonnet-5"
 
 
 class StartShadowEvalRequest(BaseModel):
-    """Start duplicating a key's traffic for blind comparison against an auto-router."""
+    """Start duplicating one or more keys' traffic for blind comparison against an auto-router."""
 
-    api_key_id: str = Field(
+    api_key_ids: tuple[str, ...] = Field(
+        min_length=1,
+        max_length=100,
         description=(
-            "The hashed virtual key whose traffic will be shadowed. Shadow evaluation runs ONLY on this "
-            "key's traffic; requests made with any other key are not sampled."
-        )
+            "The hashed virtual keys whose traffic will be shadowed. Shadow evaluation runs ONLY on these "
+            "keys' traffic; requests made with any other key are not sampled. Each key carries its own "
+            "max_turns budget, so one key exhausting its budget leaves the others sampling. At most 100 "
+            "keys per job, which also bounds every read the job's endpoints make."
+        ),
     )
     router_name: str = Field(description="The auto-router under evaluation, in either direction")
     direction: ShadowEvalDirection = Field(
@@ -204,8 +208,9 @@ class StartShadowEvalRequest(BaseModel):
         ge=1,
         le=2000,
         description=(
-            "Sample budget: the job judges at most this many turns, then completes. This is also the spend "
-            "bound; expected judge cost is roughly max_turns times one judge call"
+            "Per-key sample budget: the job judges at most this many turns of EACH scoped key's traffic, "
+            "so a job over N keys judges at most N times max_turns turns. This is also the spend bound; "
+            "expected judge cost is roughly that turn ceiling times one judge call"
         ),
     )
 
@@ -214,6 +219,12 @@ class StartShadowEvalRequest(BaseModel):
     def _round_percentage(cls, value: float) -> float:
         return round(value, 2)
 
+    @field_validator("api_key_ids")
+    @classmethod
+    def _dedupe_keys(cls, value: tuple[str, ...]) -> tuple[str, ...]:
+        """A key named twice would collide with itself on the one-active-per-(key, direction) index."""
+        return tuple(dict.fromkeys(value))
+
     @model_validator(mode="after")
     def _baseline_model_matches_direction(self) -> "StartShadowEvalRequest":
         if self.direction == "reverse" and self.baseline_model is None:
@@ -251,24 +262,46 @@ class ShadowEvalResult(BaseModel):
     by_tier: tuple[ShadowEvalSlice, ...]
     by_current_model: tuple[ShadowEvalSlice, ...] = Field(
         description=(
-            "Sliced by the model that served the real arm: the key's incumbent models in forward mode, "
+            "Sliced by the model that served the real arm: the keys' incumbent models in forward mode, "
             "and in reverse the models the router itself picked"
         )
     )
+    by_key: tuple[ShadowEvalSlice, ...] = Field(
+        description=(
+            "One slice per scoped key that has judged verdicts, grouped on the raw key hash. Keys the job "
+            "scopes but has not judged a turn for yet are absent rather than reported as zero"
+        ),
+    )
     overall_shadow_win_rate_pct: float
     overall_tie_rate_pct: float
 
 
-class ShadowEvalJobResponse(BaseModel):
-    """A shadow-eval job. Validates directly from the prisma record (job_id reads the
-    row's id); status is derived from stopped_at and ends_at, never stored, so no writer
-    anywhere can produce an inconsistent one. Aggregate fields are populated by the
-    detail endpoint only and stay None on list responses."""
+class ShadowEvalJobKeyResponse(BaseModel):
+    """One key a job shadows, with its own budget and stop state."""
 
-    model_config = ConfigDict(from_attributes=True, populate_by_name=True)
+    api_key_id: str = Field(description="The hashed virtual key whose traffic this entry scopes")
+    max_turns: int = Field(description="This key's own sample budget, independent of its siblings'")
+    stopped_at: datetime | None = Field(
+        default=None,
+        description=(
+            "When this key's slot was stamped free, whether its own budget ran out, the window closed, "
+            "or an operator stopped the job; status is derived, so a spent budget reads completed even "
+            "while this is still unset"
+        ),
+    )
+    attempt_count: int | None = Field(
+        default=None,
+        description=(
+            "This key's sampled attempts so far, judged and errored alike, the same count the sampler "
+            "budgets against max_turns; populated on list and detail responses. Frozen at stopped_at "
+            "once the key is stamped, so in-flight attempts landing after a stop never reclassify it"
+        ),
+    )
+
+    @property
+    def budget_spent(self) -> bool:
+        return self.attempt_count is not None and self.attempt_count >= self.max_turns
 
-    job_id: str = Field(validation_alias=AliasChoices("id", "job_id"))
-    api_key_id: str = Field(description="The hashed virtual key whose traffic this job evaluates, and only that key's")
     key_alias: str | None = Field(
         default=None,
         description="Alias of the shadowed key, resolved from the key row at read time; None when unset or deleted",
@@ -277,15 +310,34 @@ class ShadowEvalJobResponse(BaseModel):
         default=None,
         description="Masked display name (sk-...) of the shadowed key, resolved at read time like key_alias",
     )
+
+
+class ShadowEvalJobResponse(BaseModel):
+    """A shadow-eval job over one or more keys, each with its own budget and stop state;
+    status is derived from stopped_by, the keys' stop and budget state, and ends_at,
+    never stored, so no writer anywhere can produce an inconsistent one. Aggregate
+    fields are populated by the detail endpoint only and stay None on list responses."""
+
+    job_id: str
+    keys: tuple[ShadowEvalJobKeyResponse, ...] = Field(
+        min_length=1,
+        description="The keys whose traffic this job evaluates, and only those keys', each with its own budget",
+    )
     router_name: str
     direction: ShadowEvalDirection = "forward"
     baseline_model: str | None = None
     judge_model: str
     shadow_percentage: float
-    max_turns: int
     created_at: datetime
     ends_at: datetime
-    stopped_at: datetime | None = None
+    stopped_by: str | None = Field(
+        default=None,
+        description=(
+            "The operator who stopped the job early, recorded by the stop endpoint; 'unknown' backfilled "
+            "by migration for jobs that displayed stopped when the column arrived; None when the job "
+            "ended on its own. Its presence is what makes a job read stopped rather than completed"
+        ),
+    )
 
     judged_count: int | None = Field(default=None, description="Verdicts recorded; detail endpoint only")
     error_count: int | None = Field(default=None, description="Sampled attempts that errored; detail endpoint only")
@@ -296,12 +348,19 @@ class ShadowEvalJobResponse(BaseModel):
     @computed_field
     @property
     def status(self) -> ShadowEvalStatus:
-        """A job whose window has passed reads completed even if a later sweep stamped
-        stopped_at; stopped means sampling ended before the window did."""
+        """Three recorded facts, no history-guessing: a stop is stopped_by (the migration
+        backfills it for every job that displayed stopped when the column arrived, so the
+        pre-column population is closed), completion is the window passing or every key
+        spending its budget, and anything else is running. The all-keys-stamped fallback
+        covers only stops written by pre-column pods during a rolling deploy."""
+        if self.stopped_by is not None:
+            return "stopped"
         if datetime.now(timezone.utc) >= (
             self.ends_at if self.ends_at.tzinfo else self.ends_at.replace(tzinfo=timezone.utc)
         ):
             return "completed"
-        if self.stopped_at is not None:
+        if all(key.budget_spent for key in self.keys):
+            return "completed"
+        if all(key.stopped_at is not None for key in self.keys):
             return "stopped"
         return "running"
diff --git a/schema.prisma b/schema.prisma
index 52fb447157b..f79e2bb0c18 100644
--- a/schema.prisma
+++ b/schema.prisma
@@ -1467,28 +1467,38 @@ model LiteLLM_AutoRouterSession {
   @@index([last_turn_at], map: "idx_autorouter_session_last_turn")
 }
 
-// Shadow eval: evaluation of an auto-router against a key's live traffic, in either
-// direction. forward duplicates the requests the key did not route through the router
-// through it, answering whether the key should adopt it; reverse duplicates the requests
-// the router did serve against a fixed baseline model, answering whether a key already on
-// it still benefits. Either way a sampled slice runs in a detached task and an LLM judge
-// compares real vs shadow responses blind. The job row is immutable config plus
-// stopped_at; every count, status, and spend figure is derived from the append-only
-// attempt rows, so nothing can disagree across pods or stop races.
+// Shadow eval: evaluation of an auto-router against one or more keys' live traffic, in
+// either direction. forward duplicates the requests the keys did not route through the
+// router through it, answering whether they should adopt it; reverse duplicates the
+// requests the router did serve against a fixed baseline model, answering whether a key
+// already on it still benefits. Either way a sampled slice runs in a detached task and an
+// LLM judge compares real vs shadow responses blind. Each row is ONE key's leg of a job:
+// immutable config plus that key's own turn budget and stop state, so one key exhausting
+// its budget never ends a sibling's sampling. A job is the set of legs sharing group_id
+// (the id the API reports), written together by one atomic create_many with identical
+// config; single-key jobs predating group_id were backfilled group_id = id. "One active
+// job per (key, direction)" is a partial unique index on (api_key_id, direction) WHERE
+// stopped_at IS NULL, expressed only in the migration because schema.prisma cannot state
+// partial indexes; it is what makes a concurrent start on another pod race-safe rather
+// than read-then-create. Every count, status, and spend figure is derived from the
+// append-only attempt rows, so nothing can disagree across pods or stop races.
 model LiteLLM_ShadowEvalJob {
   id                String   @id @default(cuid())
-  api_key_id        String   // hashed virtual key whose traffic is shadowed
+  group_id          String   // legs of one job share this; the API's job id
+  api_key_id        String   // hashed virtual key whose traffic this leg shadows
   router_name       String   // the auto-router under evaluation, in either direction
   direction         String   @default("forward") // forward | reverse
   baseline_model    String?  // reverse only: the fixed model the router is judged against
   judge_model       String
   shadow_percentage Float
-  max_turns         Int      // sample budget: judge at most this many turns
+  max_turns         Int      // this key's sample budget: judge at most this many turns
   created_at        DateTime @default(now())
   created_by        String?
   ends_at           DateTime
   stopped_at        DateTime?
+  stopped_by        String?  // operator who stopped it early; null when it ended on its own
 
+  @@index([group_id])
   @@index([api_key_id])
   @@index([created_at])
 }
diff --git a/tests/test_litellm/proxy/management_endpoints/test_auto_router_endpoints.py b/tests/test_litellm/proxy/management_endpoints/test_auto_router_endpoints.py
index 77149457e82..3fd023552f5 100644
--- a/tests/test_litellm/proxy/management_endpoints/test_auto_router_endpoints.py
+++ b/tests/test_litellm/proxy/management_endpoints/test_auto_router_endpoints.py
@@ -4,6 +4,7 @@ Unit tests for auto router management endpoints
 
 import os
 import sys
+from pathlib import Path
 
 import pytest
 from fastapi import HTTPException
@@ -325,9 +326,7 @@ class TestAutoRouterBenchmarks:
         from litellm.proxy.management_endpoints.auto_router_endpoints import _benchmark_totals
 
         totals = _benchmark_totals(self.ROW)
-        bucket_hits = (
-            totals.cache.same_model.hits + totals.cache.first_visit.hits + totals.cache.return_to_tier.hits
-        )
+        bucket_hits = totals.cache.same_model.hits + totals.cache.first_visit.hits + totals.cache.return_to_tier.hits
         assert bucket_hits == 27
         assert totals.cache.hit_rate_pct == pytest.approx(100.0 * 28 / 38, abs=0.1)
 
@@ -490,7 +489,7 @@ from litellm.proxy.management_endpoints.auto_router_endpoints import (
     start_shadow_eval,
     stop_shadow_eval_job,
 )
-from litellm.types.management_endpoints.auto_router_endpoints import ShadowEvalJobResponse, StartShadowEvalRequest
+from litellm.types.management_endpoints.auto_router_endpoints import StartShadowEvalRequest
 
 VIEWER = UserAPIKeyAuth(user_role=LitellmUserRoles.PROXY_ADMIN_VIEW_ONLY, api_key="sk-view", user_id="viewer")
 NON_ADMIN = UserAPIKeyAuth(user_role=LitellmUserRoles.INTERNAL_USER, api_key="sk-user", user_id="user")
@@ -507,19 +506,23 @@ def _shadow_router() -> MagicMock:
     return router
 
 
-def _job_record(**overrides: object) -> MagicMock:
-    """Spec'd like a real prisma row: only the table's columns exist as attributes, so
-    from_attributes validation falls back to model defaults for everything else."""
+def _leg_record(**overrides: object) -> MagicMock:
+    """Spec'd like a real prisma row: only the table's columns exist as attributes. One
+    row is one key's leg of a job; legs sharing group_id are one job."""
     defaults = {
-        "id": "job-1",
+        "id": "leg-1",
+        "group_id": "job-1",
         "api_key_id": "key-hash",
         "router_name": "my-router",
+        "direction": "forward",
+        "baseline_model": None,
         "judge_model": "anthropic/claude-sonnet-5",
         "shadow_percentage": 10.0,
         "max_turns": 200,
         "created_at": datetime(2026, 8, 11, tzinfo=timezone.utc),
         "ends_at": datetime.now(timezone.utc) + timedelta(days=7),
         "stopped_at": None,
+        "stopped_by": None,
     }
     fields = {**defaults, **overrides}
     record = MagicMock(spec=list(fields))
@@ -538,23 +541,90 @@ def _key_record(
     return record
 
 
-def _shadow_prisma(active_job=None, agg_rows=None) -> MagicMock:
+def _shadow_prisma(legs=(), agg_rows=None, by_leg_rows=None, known_keys=("key-hash", "key-hash-2")) -> MagicMock:
+    """The job-table fake honours the filters it is handed, so a read that forgets
+    stopped_at sees rows the partial index would have released, one that forgets
+    direction sees the opposite-direction legs a key may hold at the same time, and a
+    group read that matched on a leg id would come back empty."""
     prisma = MagicMock()
-    prisma.db.litellm_verificationtoken.find_unique = AsyncMock(return_value=_key_record())
-    prisma.db.litellm_verificationtoken.find_many = AsyncMock(return_value=[_key_record()])
-    prisma.db.execute_raw = AsyncMock(return_value=0)
-    prisma.db.litellm_shadowevaljob.find_first = AsyncMock(return_value=active_job)
-    prisma.db.litellm_shadowevaljob.find_unique = AsyncMock(return_value=None)
-    prisma.db.litellm_shadowevaljob.find_many = AsyncMock(return_value=[])
-    prisma.db.litellm_shadowevaljob.create = AsyncMock(return_value=_job_record())
-    prisma.db.litellm_shadowevaljob.update = AsyncMock(
-        return_value=_job_record(stopped_at=datetime.now(timezone.utc))
-    )
+    prisma.db.litellm_verificationtoken.find_many = AsyncMock(return_value=[_key_record(token) for token in known_keys])
+    async def execute_raw(sql: str, *params: object):
+        if "SET stopped_by" in sql:
+            group = [row for row in stored if row.group_id == params[0]]
+            counts = {row["job_id"]: row["attempt_count"] for row in prisma.attempt_rows}
+            sampling = any(row.stopped_at is None and counts.get(row.id, 0) < row.max_turns for row in group)
+            window_open = bool(group) and group[0].ends_at > datetime.now(timezone.utc)
+            claimable = [row for row in group if row.stopped_by is None]
+            if not (claimable and sampling and window_open):
+                return 0
+            for row in claimable:
+                row.stopped_by = params[1]
+                if row.stopped_at is None:
+                    row.stopped_at = datetime.fromisoformat(str(params[2])).replace(tzinfo=timezone.utc)
+            return len(claimable)
+        return 0
+
+    prisma.db.execute_raw = AsyncMock(side_effect=execute_raw)
+    stored = legs if isinstance(legs, list) else list(legs)
+
+    async def find_many_legs(where=None, **_: object):
+        current = list(stored)
+        w = dict(where or {})
+        if "api_key_id" in w:
+            wanted = w["api_key_id"]["in"] if isinstance(w["api_key_id"], dict) else [w["api_key_id"]]
+            current = [row for row in current if row.api_key_id in wanted]
+        if "direction" in w:
+            current = [row for row in current if row.direction == w["direction"]]
+        if "stopped_at" in w:
+            current = [row for row in current if row.stopped_at is w["stopped_at"]]
+        if "group_id" in w:
+            wanted = w["group_id"]["in"] if isinstance(w["group_id"], dict) else [w["group_id"]]
+            current = [row for row in current if row.group_id in wanted]
+        return current
+
+    def newest_groups(rows, limit):
+        latest: dict = {}
+        for row in rows:
+            if row.group_id not in latest or row.created_at > latest[row.group_id]:
+                latest[row.group_id] = row.created_at
+        ordered = sorted(latest, key=lambda group_id: latest[group_id], reverse=True)
+        return ordered[: int(limit)]
+
+    def leg_dict(row):
+        fields = (
+            "id",
+            "group_id",
+            "api_key_id",
+            "router_name",
+            "direction",
+            "baseline_model",
+            "judge_model",
+            "shadow_percentage",
+            "max_turns",
+            "created_at",
+            "ends_at",
+            "stopped_at",
+            "stopped_by",
+        )
+        return {field: getattr(row, field) for field in fields}
+
+    prisma.db.litellm_shadowevaljob.find_many = AsyncMock(side_effect=find_many_legs)
+    prisma.db.litellm_shadowevaljob.create_many = AsyncMock(return_value=1)
+    prisma.db.litellm_shadowevaljob.update_many = AsyncMock(return_value=1)
     prisma.db.litellm_shadowevalattempt.find_first = AsyncMock(return_value=None)
+    prisma.attempt_rows = []
 
     async def query_raw(sql: str, *params: object):
+        if "AS attempt_count" in sql:
+            return prisma.attempt_rows
+        if "GROUP BY group_id" in sql:
+            scoped = [row for row in stored if "api_key_id = $2" not in sql or row.api_key_id == params[1]]
+            keep = set(newest_groups(scoped, params[0]))
+            return [leg_dict(row) for row in stored if row.group_id in keep]
         if "FILTER (WHERE outcome != 'error')::int AS judged_count" in sql:
             return [{"judged_count": 10, "error_count": 2, "judge_spend": 0.031}]
+        if "SELECT job_id AS grp" in sql:
+            return by_leg_rows if by_leg_rows is not None else []
         return agg_rows if agg_rows is not None else []
 
     prisma.db.query_raw = AsyncMock(side_effect=query_raw)
@@ -563,7 +633,7 @@ def _shadow_prisma(active_job=None, agg_rows=None) -> MagicMock:
 
 def _start_request(**overrides: object) -> StartShadowEvalRequest:
     payload = {
-        "api_key_id": "key-hash",
+        "api_key_ids": ("key-hash",),
         "router_name": "my-router",
         "shadow_percentage": 10.0,
         "judge_model": "anthropic/claude-sonnet-5",
@@ -575,44 +645,55 @@ def _start_request(**overrides: object) -> StartShadowEvalRequest:
 
 
 @pytest.mark.asyncio
-async def test_start_shadow_eval_creates_job_and_frees_expired_or_exhausted_ones(monkeypatch: pytest.MonkeyPatch):
-    """Expiry and turn-budget exhaustion both end sampling on their own; either must
-    release the key's slot in the active-job index so a new eval can start."""
+async def test_start_shadow_eval_writes_one_leg_per_key_in_one_statement(monkeypatch: pytest.MonkeyPatch):
+    """N keys become N sibling rows sharing group_id and identical config, written by a
+    single create_many so a unique-index loser rolls back the whole claim, and expiry or
+    budget exhaustion frees every requested key's slot first."""
     import litellm.proxy.proxy_server as proxy_server
 
     prisma = _shadow_prisma()
     monkeypatch.setattr(proxy_server, "prisma_client", prisma)
     monkeypatch.setattr(proxy_server, "llm_router", _shadow_router())
 
-    response = await start_shadow_eval(_start_request(), ADMIN)
+    response = await start_shadow_eval(_start_request(api_key_ids=("key-hash", "key-hash-2")), ADMIN)
 
-    assert response.status == "running"
-    assert response.max_turns == 200
-    assert response.judged_count is None
-    sweep_sql, sweep_key = prisma.db.execute_raw.call_args.args
+    sweep_sql, sweep_keys = prisma.db.execute_raw.call_args.args
     assert "stopped_at IS NULL" in sweep_sql
-    assert "ends_at <= NOW()" in sweep_sql
+    assert "j.ends_at <= (NOW() AT TIME ZONE 'utc')" in sweep_sql
+    assert "SET stopped_at = (NOW() AT TIME ZONE 'utc')" in sweep_sql
     assert ">= j.max_turns" in sweep_sql
-    assert sweep_key == "key-hash"
-    create_data = prisma.db.litellm_shadowevaljob.create.call_args.kwargs["data"]
-    assert create_data["api_key_id"] == "key-hash"
-    assert create_data["created_by"] == "admin"
-    assert "status" not in create_data
+    assert "j.api_key_id = ANY($1::text[])" in sweep_sql
+    assert sweep_keys == ["key-hash", "key-hash-2"]
+    prisma.db.litellm_shadowevaljob.create_many.assert_awaited_once()
+    rows = prisma.db.litellm_shadowevaljob.create_many.call_args.kwargs["data"]
+    assert [row["api_key_id"] for row in rows] == ["key-hash", "key-hash-2"]
+    assert len({frozenset((k, v) for k, v in row.items() if k != "api_key_id") for row in rows}) == 1
+    assert len({row["group_id"] for row in rows}) == 1
+    assert all(row["max_turns"] == 200 and row["created_by"] == "admin" for row in rows)
+    assert all("status" not in row and "id" not in row for row in rows)
+    assert response.job_id == rows[0]["group_id"]
+    assert response.status == "running"
+    assert response.judged_count is None
+    assert [(key.api_key_id, key.max_turns, key.key_alias) for key in response.keys] == [
+        ("key-hash", 200, "prod-alpha"),
+        ("key-hash-2", 200, "prod-alpha"),
+    ]
 
 
 @pytest.mark.asyncio
 @pytest.mark.parametrize(
-    "caller,request_overrides,active,expected_status",
+    "caller,request_overrides,claimed,expected_status",
     [
-        (NON_ADMIN, {}, None, 403),
-        (VIEWER, {}, None, 403),
-        (ADMIN, {"router_name": "not-a-router"}, None, 400),
-        (ADMIN, {"judge_model": "not/a real model!"}, None, 400),
-        (ADMIN, {"judge_model": "my-router"}, None, 400),
-        (ADMIN, {}, "active", 409),
-        (ADMIN, {"direction": "reverse", "baseline_model": "my-router"}, None, 400),
-        (ADMIN, {"direction": "reverse", "baseline_model": "not/a real model!"}, None, 400),
-        (ADMIN, {"direction": "reverse", "baseline_model": "openai/gpt-4o", "router_name": "not-a-router"}, None, 400),
+        (NON_ADMIN, {}, (), 403),
+        (VIEWER, {}, (), 403),
+        (ADMIN, {"router_name": "not-a-router"}, (), 400),
+        (ADMIN, {"judge_model": "not/a real model!"}, (), 400),
+        (ADMIN, {"judge_model": "my-router"}, (), 400),
+        (ADMIN, {}, ("key-hash",), 409),
+        (ADMIN, {"api_key_ids": ("key-hash", "key-hash-2")}, ("key-hash-2",), 409),
+        (ADMIN, {"direction": "reverse", "baseline_model": "my-router"}, (), 400),
+        (ADMIN, {"direction": "reverse", "baseline_model": "not/a real model!"}, (), 400),
+        (ADMIN, {"direction": "reverse", "baseline_model": "openai/gpt-4o", "router_name": "not-a-router"}, (), 400),
     ],
     ids=[
         "non-admin",
@@ -621,23 +702,143 @@ async def test_start_shadow_eval_creates_job_and_frees_expired_or_exhausted_ones
         "unresolvable-judge",
         "router-as-judge",
         "already-active",
+        "one-of-several-keys-already-active",
         "router-as-baseline",
         "unresolvable-baseline",
         "reverse-still-needs-an-auto-router",
     ],
 )
 async def test_start_shadow_eval_rejections(
-    monkeypatch: pytest.MonkeyPatch, caller, request_overrides, active, expected_status
+    monkeypatch: pytest.MonkeyPatch, caller, request_overrides, claimed, expected_status
 ):
     import litellm.proxy.proxy_server as proxy_server
 
-    prisma = _shadow_prisma(active_job=_job_record() if active else None)
+    prisma = _shadow_prisma(legs=[_leg_record(id=f"leg-{key}", group_id="job-7", api_key_id=key) for key in claimed])
     monkeypatch.setattr(proxy_server, "prisma_client", prisma)
     monkeypatch.setattr(proxy_server, "llm_router", _shadow_router())
 
     with pytest.raises(HTTPException) as exc:
         await start_shadow_eval(_start_request(**request_overrides), caller)
     assert exc.value.status_code == expected_status
+    prisma.db.litellm_shadowevaljob.create_many.assert_not_called()
+
+
+@pytest.mark.asyncio
+async def test_start_shadow_eval_names_the_busy_key_and_its_job(monkeypatch: pytest.MonkeyPatch):
+    """A key busy elsewhere blocks the whole start rather than being silently dropped from
+    it, and the 409 names which key and which job so the caller can stop or drop it."""
+    import litellm.proxy.proxy_server as proxy_server
+
+    prisma = _shadow_prisma(legs=[_leg_record(id="leg-b", group_id="job-7", api_key_id="key-hash-2")])
+    monkeypatch.setattr(proxy_server, "prisma_client", prisma)
+    monkeypatch.setattr(proxy_server, "llm_router", _shadow_router())
+
+    with pytest.raises(HTTPException) as exc:
+        await start_shadow_eval(_start_request(api_key_ids=("key-hash", "key-hash-2")), ADMIN)
+    assert exc.value.status_code == 409
+    assert "key-hash-2 (job job-7)" in exc.value.detail
+
+
+@pytest.mark.asyncio
+async def test_start_shadow_eval_reuses_a_key_whose_previous_job_already_stopped(monkeypatch: pytest.MonkeyPatch):
+    """The claim is held by unstopped legs only, matching the partial unique index. A read
+    that forgets that would strand every key that has ever finished a job."""
+    import litellm.proxy.proxy_server as proxy_server
+
+    prisma = _shadow_prisma(legs=[_leg_record(group_id="job-7", stopped_at=datetime.now(timezone.utc))])
+    monkeypatch.setattr(proxy_server, "prisma_client", prisma)
+    monkeypatch.setattr(proxy_server, "llm_router", _shadow_router())
+
+    job = await start_shadow_eval(_start_request(), ADMIN)
+
+    assert job.status == "running"
+    prisma.db.litellm_shadowevaljob.create_many.assert_awaited_once()
+
+
+@pytest.mark.asyncio
+async def test_start_shadow_eval_reverse_records_its_arms_and_holds_its_own_slot(monkeypatch: pytest.MonkeyPatch):
+    """The two directions ask opposite questions of the same key, so a forward job holding
+    the slot must not block a reverse one. The second reverse start still 409s."""
+    import litellm.proxy.proxy_server as proxy_server
+
+    legs = [_leg_record(group_id="job-fwd")]
+    prisma = _shadow_prisma(legs=legs)
+    monkeypatch.setattr(proxy_server, "prisma_client", prisma)
+    monkeypatch.setattr(proxy_server, "llm_router", _shadow_router())
+
+    reverse = _start_request(direction="reverse", baseline_model="openai/gpt-4o")
+    response = await start_shadow_eval(reverse, ADMIN)
+
+    assert (response.direction, response.baseline_model) == ("reverse", "openai/gpt-4o")
+    rows = prisma.db.litellm_shadowevaljob.create_many.call_args.kwargs["data"]
+    assert rows[0]["direction"] == "reverse"
+    assert rows[0]["baseline_model"] == "openai/gpt-4o"
+
+    legs.append(_leg_record(id="leg-2", group_id="job-rev", direction="reverse"))
+    with pytest.raises(HTTPException) as exc:
+        await start_shadow_eval(reverse, ADMIN)
+    assert exc.value.status_code == 409
+
+
+@pytest.mark.asyncio
+async def test_start_shadow_eval_forward_leaves_the_baseline_column_empty(monkeypatch: pytest.MonkeyPatch):
+    import litellm.proxy.proxy_server as proxy_server
+
+    prisma = _shadow_prisma()
+    monkeypatch.setattr(proxy_server, "prisma_client", prisma)
+    monkeypatch.setattr(proxy_server, "llm_router", _shadow_router())
+
+    await start_shadow_eval(_start_request(), ADMIN)
+
+    rows = prisma.db.litellm_shadowevaljob.create_many.call_args.kwargs["data"]
+    assert rows[0]["direction"] == "forward"
+    assert rows[0]["baseline_model"] is None
+
+
+@pytest.mark.asyncio
+async def test_start_shadow_eval_rejects_keys_this_proxy_does_not_know(monkeypatch: pytest.MonkeyPatch):
+    """A typo'd api_key_id would otherwise create a leg no traffic can ever match. Every
+    unknown key is named at once, so a caller passing several fixes them in one round."""
+    import litellm.proxy.proxy_server as proxy_server
+
+    prisma = _shadow_prisma(known_keys=("key-hash",))
+    monkeypatch.setattr(proxy_server, "prisma_client", prisma)
+    monkeypatch.setattr(proxy_server, "llm_router", _shadow_router())
+
+    with pytest.raises(HTTPException) as exc:
+        await start_shadow_eval(_start_request(api_key_ids=("key-hash", "typo-a", "typo-b")), ADMIN)
+    assert exc.value.status_code == 400
+    assert "typo-a, typo-b" in exc.value.detail
+    assert "key-hash," not in exc.value.detail
+    prisma.db.litellm_shadowevaljob.create_many.assert_not_called()
+
+
+def test_start_shadow_eval_request_dedupes_and_bounds_the_key_set():
+    """A key named twice would collide with itself on the one-active-per-key index, a job
+    scoping no key samples nothing, and the key-count cap bounds every downstream read."""
+    assert _start_request(api_key_ids=("a", "b", "a")).api_key_ids == ("a", "b")
+    assert len(_start_request(api_key_ids=tuple(f"k{i}" for i in range(100))).api_key_ids) == 100
+    with pytest.raises(ValidationError):
+        _start_request(api_key_ids=())
+    with pytest.raises(ValidationError):
+        _start_request(api_key_ids=tuple(f"k{i}" for i in range(101)))
+
+
+@pytest.mark.asyncio
+async def test_start_shadow_eval_concurrent_unique_violation_is_a_409(monkeypatch: pytest.MonkeyPatch):
+    import litellm.proxy.proxy_server as proxy_server
+    from prisma.errors import UniqueViolationError
+
+    prisma = _shadow_prisma()
+    prisma.db.litellm_shadowevaljob.create_many = AsyncMock(
+        side_effect=UniqueViolationError(MagicMock(message="unique constraint"))
+    )
+    monkeypatch.setattr(proxy_server, "prisma_client", prisma)
+    monkeypatch.setattr(proxy_server, "llm_router", _shadow_router())
+
+    with pytest.raises(HTTPException) as exc:
+        await start_shadow_eval(_start_request(), ADMIN)
+    assert exc.value.status_code == 409
 
 
 @pytest.mark.parametrize(
@@ -657,97 +858,25 @@ def test_start_request_pins_baseline_model_to_reverse(overrides):
 
 
 @pytest.mark.asyncio
-async def test_start_shadow_eval_reverse_records_its_arms_and_holds_its_own_slot(monkeypatch: pytest.MonkeyPatch):
-    """The two directions ask opposite questions of the same key, so a forward job holding
-    the slot must not block a reverse one. The second reverse start still 409s."""
-    import litellm.proxy.proxy_server as proxy_server
-
-    prisma = _shadow_prisma()
-    active = {"forward": _job_record()}
-    prisma.db.litellm_shadowevaljob.find_first = AsyncMock(
-        side_effect=lambda where, **_: active.get(str(where.get("direction")))
-    )
-    prisma.db.litellm_shadowevaljob.create = AsyncMock(
-        return_value=_job_record(direction="reverse", baseline_model="openai/gpt-4o")
-    )
-    monkeypatch.setattr(proxy_server, "prisma_client", prisma)
-    monkeypatch.setattr(proxy_server, "llm_router", _shadow_router())
-
-    reverse = _start_request(direction="reverse", baseline_model="openai/gpt-4o")
-    response = await start_shadow_eval(reverse, ADMIN)
-
-    assert (response.direction, response.baseline_model) == ("reverse", "openai/gpt-4o")
-    create_data = prisma.db.litellm_shadowevaljob.create.call_args.kwargs["data"]
-    assert create_data["direction"] == "reverse"
-    assert create_data["baseline_model"] == "openai/gpt-4o"
-
-    active["reverse"] = _job_record(id="job-2", direction="reverse")
-    with pytest.raises(HTTPException) as exc:
-        await start_shadow_eval(reverse, ADMIN)
-    assert exc.value.status_code == 409
-
-
-@pytest.mark.asyncio
-async def test_start_shadow_eval_forward_leaves_the_baseline_column_empty(monkeypatch: pytest.MonkeyPatch):
-    import litellm.proxy.proxy_server as proxy_server
-
-    prisma = _shadow_prisma()
-    monkeypatch.setattr(proxy_server, "prisma_client", prisma)
-    monkeypatch.setattr(proxy_server, "llm_router", _shadow_router())
-
-    await start_shadow_eval(_start_request(), ADMIN)
-
-    create_data = prisma.db.litellm_shadowevaljob.create.call_args.kwargs["data"]
-    assert create_data["direction"] == "forward"
-    assert create_data["baseline_model"] is None
-
-
-@pytest.mark.asyncio
-async def test_start_shadow_eval_rejects_a_key_this_proxy_does_not_know(monkeypatch: pytest.MonkeyPatch):
-    """A typo'd api_key_id would otherwise create a job no traffic can ever match."""
-    import litellm.proxy.proxy_server as proxy_server
-
-    prisma = _shadow_prisma()
-    prisma.db.litellm_verificationtoken.find_unique = AsyncMock(return_value=None)
-    monkeypatch.setattr(proxy_server, "prisma_client", prisma)
-    monkeypatch.setattr(proxy_server, "llm_router", _shadow_router())
-
-    with pytest.raises(HTTPException) as exc:
-        await start_shadow_eval(_start_request(), ADMIN)
-    assert exc.value.status_code == 400
-    assert "not a key on this proxy" in exc.value.detail
-
-
-@pytest.mark.asyncio
-async def test_start_shadow_eval_concurrent_unique_violation_is_a_409(monkeypatch: pytest.MonkeyPatch):
-    import litellm.proxy.proxy_server as proxy_server
-    from prisma.errors import UniqueViolationError
-
-    prisma = _shadow_prisma()
-    prisma.db.litellm_shadowevaljob.create = AsyncMock(
-        side_effect=UniqueViolationError(MagicMock(message="unique constraint"))
-    )
-    monkeypatch.setattr(proxy_server, "prisma_client", prisma)
-    monkeypatch.setattr(proxy_server, "llm_router", _shadow_router())
-
-    with pytest.raises(HTTPException) as exc:
-        await start_shadow_eval(_start_request(), ADMIN)
-    assert exc.value.status_code == 409
-
-
-@pytest.mark.asyncio
-async def test_get_shadow_eval_job_derives_counts_spend_and_stratified_results(monkeypatch: pytest.MonkeyPatch):
+async def test_get_shadow_eval_job_pools_counts_and_slices_results_per_key(monkeypatch: pytest.MonkeyPatch):
+    """One read answers for every leg: totals and stratifications aggregate over the
+    group's leg ids, and the by-key slice maps each leg id back to its key hash."""
     import litellm.proxy.proxy_server as proxy_server
 
     tier_rows = [
         {"grp": "SIMPLE", "turn_count": 8, "real_wins": 2, "shadow_wins": 4, "ties": 2, "avg_confidence": 0.8},
         {"grp": "REASONING", "turn_count": 2, "real_wins": 2, "shadow_wins": 0, "ties": 0, "avg_confidence": 0.9},
     ]
-    prisma = _shadow_prisma(agg_rows=tier_rows)
-    prisma.db.litellm_shadowevaljob.find_unique = AsyncMock(return_value=_job_record())
-    prisma.db.litellm_shadowevalattempt.find_first = AsyncMock(
-        return_value=MagicMock(error="judge call failed: boom")
+    leg_rows = [
+        {"grp": "leg-1", "turn_count": 6, "real_wins": 1, "shadow_wins": 4, "ties": 1, "avg_confidence": 0.7},
+        {"grp": "leg-2", "turn_count": 4, "real_wins": 3, "shadow_wins": 0, "ties": 1, "avg_confidence": 0.6},
+    ]
+    prisma = _shadow_prisma(
+        legs=[_leg_record(), _leg_record(id="leg-2", api_key_id="key-hash-2", max_turns=50)],
+        agg_rows=tier_rows,
+        by_leg_rows=leg_rows,
     )
+    prisma.db.litellm_shadowevalattempt.find_first = AsyncMock(return_value=MagicMock(error="judge call failed: boom"))
     monkeypatch.setattr(proxy_server, "prisma_client", prisma)
 
     response = await get_shadow_eval_job("job-1", VIEWER)
@@ -762,6 +891,13 @@ async def test_get_shadow_eval_job_derives_counts_spend_and_stratified_results(m
     assert response.results.by_tier[0].shadow_win_rate_pct == 50.0
     assert response.results.overall_shadow_win_rate_pct == 40.0
     assert response.results.overall_tie_rate_pct == 20.0
+    assert [(s.group, s.turn_count) for s in response.results.by_key] == [("key-hash", 6), ("key-hash-2", 4)]
+    assert response.results.by_key[0].shadow_win_rate_pct == 66.7
+    assert [(key.api_key_id, key.max_turns) for key in response.keys] == [("key-hash", 200), ("key-hash-2", 50)]
+    totals_args = [call.args for call in prisma.db.query_raw.await_args_list if "judged_count" in call.args[0]]
+    assert totals_args == [(totals_args[0][0], ["leg-1", "leg-2"])]
+    error_where = prisma.db.litellm_shadowevalattempt.find_first.call_args.kwargs["where"]
+    assert error_where == {"job_id": {"in": ["leg-1", "leg-2"]}, "outcome": "error"}
 
 
 @pytest.mark.asyncio
@@ -780,79 +916,326 @@ async def test_get_shadow_eval_job_404s_and_gates_on_role(monkeypatch: pytest.Mo
 
 
 @pytest.mark.asyncio
-async def test_list_shadow_eval_jobs_returns_derived_status_without_aggregates(monkeypatch: pytest.MonkeyPatch):
+async def test_list_shadow_eval_jobs_collapses_legs_into_jobs_newest_first(monkeypatch: pytest.MonkeyPatch):
+    """A job over two keys is one list entry with both keys, not two entries, and a job
+    whose keys all stopped reads stopped while a half-stopped one still runs."""
     import litellm.proxy.proxy_server as proxy_server
 
-    prisma = _shadow_prisma()
-    prisma.db.litellm_shadowevaljob.find_many = AsyncMock(
-        return_value=[
-            _job_record(),
-            _job_record(id="job-2", ends_at=datetime.now(timezone.utc) - timedelta(days=1)),
-            _job_record(id="job-3", stopped_at=datetime.now(timezone.utc)),
+    stamp = datetime.now(timezone.utc)
+    prisma = _shadow_prisma(
+        legs=[
+            _leg_record(created_at=datetime(2026, 8, 13, tzinfo=timezone.utc)),
+            _leg_record(
+                id="leg-2",
+                api_key_id="key-hash-2",
+                stopped_at=stamp,
+                created_at=datetime(2026, 8, 13, tzinfo=timezone.utc),
+            ),
+            _leg_record(
+                id="leg-3",
+                group_id="job-2",
+                stopped_at=stamp,
+                created_at=datetime(2026, 8, 12, tzinfo=timezone.utc),
+            ),
+            _leg_record(
+                id="leg-4",
+                group_id="job-3",
+                ends_at=datetime.now(timezone.utc) - timedelta(days=1),
+                created_at=datetime(2026, 8, 11, tzinfo=timezone.utc),
+            ),
         ]
     )
     monkeypatch.setattr(proxy_server, "prisma_client", prisma)
 
     jobs = await list_shadow_eval_jobs(VIEWER, api_key_id=None, limit=50)
 
-    assert [job.status for job in jobs] == ["running", "completed", "stopped"]
-    swept = ShadowEvalJobResponse.model_validate(
-        _job_record(
-            id="job-4",
-            ends_at=datetime.now(timezone.utc) - timedelta(days=1),
-            stopped_at=datetime.now(timezone.utc),
-        ),
-        from_attributes=True,
-    )
-    assert swept.status == "completed"
+    assert [(job.job_id, job.status) for job in jobs] == [
+        ("job-1", "running"),
+        ("job-2", "stopped"),
+        ("job-3", "completed"),
+    ]
+    assert [key.api_key_id for key in jobs[0].keys] == ["key-hash", "key-hash-2"]
     assert all(job.judged_count is None and job.results is None for job in jobs)
-    assert prisma.db.query_raw.await_count == 0
+    legs_sql, legs_limit = prisma.db.query_raw.await_args_list[0].args
+    assert "GROUP BY group_id ORDER BY MAX(created_at) DESC LIMIT $1::int" in legs_sql
+    assert legs_limit == 50
+    counts_sql, _ = prisma.db.query_raw.await_args_list[1].args
+    assert "AS attempt_count" in counts_sql
+    assert "j.stopped_at IS NULL OR a.created_at <= j.stopped_at" in counts_sql
+    assert prisma.db.query_raw.await_count == 2
+    prisma.db.litellm_shadowevaljob.find_many.assert_not_called()
 
 
 @pytest.mark.asyncio
-async def test_shadow_eval_responses_name_the_shadowed_key(monkeypatch: pytest.MonkeyPatch):
+async def test_list_shadow_eval_jobs_filters_to_jobs_containing_the_key(monkeypatch: pytest.MonkeyPatch):
+    """The filter matches a key anywhere in a job's key set and still returns the whole
+    job, sibling keys included."""
     import litellm.proxy.proxy_server as proxy_server
 
-    prisma = _shadow_prisma()
-    prisma.db.litellm_shadowevaljob.find_many = AsyncMock(
-        return_value=[_job_record(), _job_record(id="job-2", api_key_id="deleted-key-hash")]
+    prisma = _shadow_prisma(
+        legs=[
+            _leg_record(),
+            _leg_record(id="leg-2", api_key_id="key-hash-2"),
+            _leg_record(id="leg-3", group_id="job-2", api_key_id="key-hash-2"),
+            _leg_record(id="leg-4", group_id="job-3"),
+        ]
+    )
+    monkeypatch.setattr(proxy_server, "prisma_client", prisma)
+
+    jobs = await list_shadow_eval_jobs(VIEWER, api_key_id="key-hash-2", limit=50)
+
+    assert [job.job_id for job in jobs] == ["job-1", "job-2"]
+    assert [key.api_key_id for key in jobs[0].keys] == ["key-hash", "key-hash-2"]
+
+
+@pytest.mark.parametrize(
+    ("stopped_flags", "days_left", "expected"),
+    [
+        ((False, False), 7, "running"),
+        ((True, False), 7, "running"),
+        ((True, True), 7, "stopped"),
+        ((True, True), -1, "completed"),
+        ((False, False), -1, "completed"),
+    ],
+)
+@pytest.mark.asyncio
+async def test_job_status_runs_until_every_key_stops_and_completed_outranks_stopped(
+    monkeypatch: pytest.MonkeyPatch, stopped_flags: tuple[bool, ...], days_left: int, expected: str
+):
+    import litellm.proxy.proxy_server as proxy_server
+
+    stamp = datetime.now(timezone.utc)
+    prisma = _shadow_prisma(
+        legs=[
+            _leg_record(
+                id=f"leg-{index}",
+                api_key_id=f"key-{index}",
+                stopped_at=stamp if stopped else None,
+                ends_at=datetime.now(timezone.utc) + timedelta(days=days_left),
+            )
+            for index, stopped in enumerate(stopped_flags)
+        ]
+    )
+    monkeypatch.setattr(proxy_server, "prisma_client", prisma)
+
+    jobs = await list_shadow_eval_jobs(VIEWER, api_key_id=None, limit=50)
+
+    assert [job.status for job in jobs] == [expected]
+
+
+@pytest.mark.asyncio
+async def test_list_reads_completed_once_every_key_spends_its_budget(monkeypatch: pytest.MonkeyPatch):
+    """A job whose keys all exhausted their turn budgets stopped sampling on its own, so
+    it must read completed on the very next list, before any sweep stamps its legs; one
+    key under budget keeps the whole job running. An operator starting an unrelated eval
+    must never look like it terminated a finished one."""
+    import litellm.proxy.proxy_server as proxy_server
+
+    prisma = _shadow_prisma(
+        legs=[
+            _leg_record(max_turns=5),
+            _leg_record(id="leg-2", api_key_id="key-hash-2", max_turns=5),
+            _leg_record(id="leg-3", group_id="job-2", api_key_id="key-hash", max_turns=5),
+            _leg_record(id="leg-4", group_id="job-2", api_key_id="key-hash-2", max_turns=5),
+        ]
+    )
+    prisma.attempt_rows = [
+        {"job_id": "leg-1", "attempt_count": 5},
+        {"job_id": "leg-2", "attempt_count": 6},
+        {"job_id": "leg-3", "attempt_count": 5},
+        {"job_id": "leg-4", "attempt_count": 3},
+    ]
+    monkeypatch.setattr(proxy_server, "prisma_client", prisma)
+
+    jobs = await list_shadow_eval_jobs(VIEWER, api_key_id=None, limit=50)
+
+    by_id = {job.job_id: job for job in jobs}
+    assert by_id["job-1"].status == "completed"
+    assert all(key.stopped_at is None for key in by_id["job-1"].keys)
+    assert by_id["job-2"].status == "running"
+    assert {key.api_key_id: key.attempt_count for key in by_id["job-2"].keys} == {"key-hash": 5, "key-hash-2": 3}
+
+
+@pytest.mark.asyncio
+async def test_recorded_operator_stop_outranks_budget_arithmetic(monkeypatch: pytest.MonkeyPatch):
+    """A detached attempt can land around the stop and push the raw count past the
+    budget; the recorded stopped_by must keep the job reading stopped regardless."""
+    import litellm.proxy.proxy_server as proxy_server
+
+    stamp = datetime.now(timezone.utc)
+    prisma = _shadow_prisma(legs=[_leg_record(max_turns=5, stopped_at=stamp, stopped_by="admin")])
+    prisma.attempt_rows = [{"job_id": "leg-1", "attempt_count": 6}]
+    monkeypatch.setattr(proxy_server, "prisma_client", prisma)
+
+    jobs = await list_shadow_eval_jobs(VIEWER, api_key_id=None, limit=50)
+    assert jobs[0].status == "stopped"
+    assert jobs[0].stopped_by == "admin"
+
+    detail = await get_shadow_eval_job("job-1", VIEWER)
+    assert detail.status == "stopped"
+
+
+@pytest.mark.asyncio
+async def test_backfilled_legacy_stop_never_reads_as_completion(monkeypatch: pytest.MonkeyPatch):
+    """Jobs stopped before stopped_by existed are backfilled with 'unknown' by the
+    migration, so even one whose stray attempts crossed the budget stays stopped."""
+    import litellm.proxy.proxy_server as proxy_server
+
+    prisma = _shadow_prisma(
+        legs=[_leg_record(max_turns=5, stopped_at=datetime.now(timezone.utc), stopped_by="unknown")]
+    )
+    prisma.attempt_rows = [{"job_id": "leg-1", "attempt_count": 6}]
+    monkeypatch.setattr(proxy_server, "prisma_client", prisma)
+
+    jobs = await list_shadow_eval_jobs(VIEWER, api_key_id=None, limit=50)
+    assert jobs[0].status == "stopped"
+
+
+def test_stopped_by_migration_backfills_every_job_that_displayed_stopped():
+    """The migration must close the pre-column population: without the backfill, a
+    legacy stop whose stray attempts crossed the budget would read completed."""
+    import litellm_proxy_extras
+
+    sql = (
+        Path(litellm_proxy_extras.__file__).parent
+        / "migrations"
+        / "20260818224500_add_shadow_eval_stopped_by"
+        / "migration.sql"
+    ).read_text()
+    assert 'ADD COLUMN     "stopped_by" TEXT' in sql
+    assert "SET stopped_by = 'unknown'" in sql
+    assert "WHERE stopped_at IS NOT NULL AND ends_at > (NOW() AT TIME ZONE 'utc')" in sql
+
+
+@pytest.mark.asyncio
+async def test_stop_rejects_a_job_that_already_spent_its_budget(monkeypatch: pytest.MonkeyPatch):
+    import litellm.proxy.proxy_server as proxy_server
+
+    prisma = _shadow_prisma(legs=[_leg_record(max_turns=3)])
+    prisma.attempt_rows = [{"job_id": "leg-1", "attempt_count": 3}]
+    monkeypatch.setattr(proxy_server, "prisma_client", prisma)
+
+    with pytest.raises(HTTPException) as exhausted:
+        await stop_shadow_eval_job("job-1", ADMIN)
+    assert exhausted.value.status_code == 400
+    assert "completed" in exhausted.value.detail
+    prisma.db.litellm_shadowevaljob.update_many.assert_not_called()
+
+
+@pytest.mark.asyncio
+async def test_shadow_eval_responses_name_every_shadowed_key(monkeypatch: pytest.MonkeyPatch):
+    import litellm.proxy.proxy_server as proxy_server
+
+    prisma = _shadow_prisma(
+        legs=[_leg_record(), _leg_record(id="leg-2", api_key_id="deleted-key-hash")],
+        known_keys=("key-hash", "key-hash-2"),
     )
     monkeypatch.setattr(proxy_server, "llm_router", _shadow_router())
     monkeypatch.setattr(proxy_server, "prisma_client", prisma)
 
-    started = await start_shadow_eval(_start_request(), ADMIN)
-    assert (started.key_alias, started.key_name) == ("prod-alpha", "sk-...lpha")
-
     jobs = await list_shadow_eval_jobs(VIEWER, api_key_id=None, limit=50)
-    assert [(job.key_alias, job.key_name) for job in jobs] == [("prod-alpha", "sk-...lpha"), (None, None)]
+    assert [(key.key_alias, key.key_name) for key in jobs[0].keys] == [
+        (None, None),
+        ("prod-alpha", "sk-...lpha"),
+    ]
     batched_where = prisma.db.litellm_verificationtoken.find_many.call_args.kwargs["where"]
     assert batched_where == {"token": {"in": ["deleted-key-hash", "key-hash"]}}
 
-    prisma.db.litellm_shadowevaljob.find_unique = AsyncMock(return_value=_job_record())
     detail = await get_shadow_eval_job("job-1", VIEWER)
-    assert detail.key_alias == "prod-alpha"
+    assert [key.key_alias for key in detail.keys] == [None, "prod-alpha"]
 
 
 @pytest.mark.asyncio
-async def test_stop_shadow_eval_sets_stopped_at_and_rejects_non_running(monkeypatch: pytest.MonkeyPatch):
+async def test_stop_shadow_eval_stops_every_unstopped_leg_and_rejects_non_running(
+    monkeypatch: pytest.MonkeyPatch,
+):
+    """One stop ends sampling for the whole job, while a leg that already stopped on its
+    own budget keeps the stopped_at it earned."""
     import litellm.proxy.proxy_server as proxy_server
 
-    prisma = _shadow_prisma()
-    prisma.db.litellm_shadowevaljob.find_unique = AsyncMock(return_value=_job_record())
+    earned = datetime.now(timezone.utc) - timedelta(hours=1)
+    prisma = _shadow_prisma(legs=[_leg_record(), _leg_record(id="leg-2", api_key_id="key-hash-2", stopped_at=earned)])
     monkeypatch.setattr(proxy_server, "prisma_client", prisma)
 
     stopped = await stop_shadow_eval_job("job-1", ADMIN)
-    assert stopped.status == "stopped"
-    update = prisma.db.litellm_shadowevaljob.update.call_args.kwargs
-    assert set(update["data"]) == {"stopped_at"}
 
-    prisma.db.litellm_shadowevaljob.find_unique = AsyncMock(
-        return_value=_job_record(ends_at=datetime.now(timezone.utc) - timedelta(days=1))
-    )
+    assert stopped.status == "stopped"
+    assert stopped.stopped_by == "admin"
+    stop_sql, stop_group, stop_operator, stop_stamp = prisma.db.execute_raw.call_args.args
+    assert "SET stopped_by = $2, stopped_at = COALESCE(stopped_at, $3::timestamp)" in stop_sql
+    assert "WHERE group_id = $1 AND stopped_by IS NULL" in stop_sql
+    assert "ends_at > (NOW() AT TIME ZONE 'utc')" in stop_sql
+    assert ") < k.max_turns" in stop_sql
+    assert (stop_group, stop_operator) == ("job-1", "admin")
+    assert datetime.fromisoformat(stop_stamp).tzinfo is None
+    assert prisma.db.execute_raw.await_count == 1
+    prisma.db.litellm_shadowevaljob.update_many.assert_not_called()
+    by_key = {key.api_key_id: key.stopped_at for key in stopped.keys}
+    assert by_key["key-hash-2"] == earned
+    assert by_key["key-hash"] is not None and by_key["key-hash"] != earned
+
+    done_leg = _leg_record(ends_at=datetime.now(timezone.utc) - timedelta(days=1))
+    prisma_done = _shadow_prisma(legs=[done_leg])
+    monkeypatch.setattr(proxy_server, "prisma_client", prisma_done)
     with pytest.raises(HTTPException) as exc:
         await stop_shadow_eval_job("job-1", ADMIN)
     assert exc.value.status_code == 400
+    assert "already completed" in exc.value.detail
+    assert done_leg.stopped_by is None
 
     with pytest.raises(HTTPException) as forbidden:
         await stop_shadow_eval_job("job-1", VIEWER)
     assert forbidden.value.status_code == 403
+
+
+def test_every_shadow_eval_sql_constant_speaks_naive_utc():
+    """The tables store naive UTC wall time (prisma's convention), so SQL-side time must be
+    NOW() AT TIME ZONE 'utc' and python-side params must cast ::timestamp; a bare NOW() or a
+    timestamptz cast writes session-local wall time into the naive column and skews every
+    comparison against prisma-written stamps."""
+    import litellm.proxy.management_endpoints.auto_router_endpoints as module
+
+    sql_constants = {name: value for name, value in vars(module).items() if name.endswith("_SQL")}
+    assert sql_constants
+    for name, sql in sql_constants.items():
+        assert "::timestamptz" not in sql, name
+        for occurrence in sql.split("NOW()")[1:]:
+            assert occurrence.startswith(" AT TIME ZONE 'utc'"), name
+
+
+@pytest.mark.asyncio
+async def test_a_stop_racing_the_last_budgeted_attempt_reports_completed_not_stopped(
+    monkeypatch: pytest.MonkeyPatch,
+):
+    """The statement claims the job only while a leg still samples, so a stop landing in
+    the same instant the budget spends records nothing and the job keeps reading
+    completed; stamping it would misreport a self-ended job as operator-stopped forever."""
+    import litellm.proxy.proxy_server as proxy_server
+
+    prisma = _shadow_prisma(legs=[_leg_record(max_turns=2)])
+    prisma.attempt_rows = [{"job_id": "leg-1", "attempt_count": 2}]
+    monkeypatch.setattr(proxy_server, "prisma_client", prisma)
+
+    with pytest.raises(HTTPException) as exc:
+        await stop_shadow_eval_job("job-1", ADMIN)
+    assert exc.value.status_code == 400
+    assert "already completed" in exc.value.detail
+    assert prisma.db.litellm_shadowevaljob.find_many.await_args.kwargs["where"] == {"group_id": "job-1"}
+
+
+@pytest.mark.asyncio
+async def test_two_racing_stops_produce_exactly_one_winner(monkeypatch: pytest.MonkeyPatch):
+    """The statement's stopped_by IS NULL predicate lets only one racer claim rows; the
+    loser reads the stamped state and gets the same answer a late caller gets."""
+    import litellm.proxy.proxy_server as proxy_server
+
+    prisma = _shadow_prisma(legs=[_leg_record()])
+    monkeypatch.setattr(proxy_server, "prisma_client", prisma)
+
+    first = await stop_shadow_eval_job("job-1", ADMIN)
+    assert first.status == "stopped"
+
+    with pytest.raises(HTTPException) as exc:
+        await stop_shadow_eval_job("job-1", ADMIN)
+    assert exc.value.status_code == 400
+    assert "already stopped" in exc.value.detail
diff --git a/ui/litellm-dashboard/src/app/(dashboard)/cost-optimization/_components/ShadowEvalSection.test.tsx b/ui/litellm-dashboard/src/app/(dashboard)/cost-optimization/_components/ShadowEvalSection.test.tsx
index 7bb550f729e..b307e3d0f2a 100644
--- a/ui/litellm-dashboard/src/app/(dashboard)/cost-optimization/_components/ShadowEvalSection.test.tsx
+++ b/ui/litellm-dashboard/src/app/(dashboard)/cost-optimization/_components/ShadowEvalSection.test.tsx
@@ -77,7 +77,15 @@ const job = (overrides: Partial = {}): ShadowEvalJob => ({
   baseline_model: null,
   judge_model: "anthropic/claude-sonnet-5",
   shadow_percentage: 10,
-  max_turns: 200,
+  keys: [
+    {
+      api_key_id: "hashed-key-abc",
+      max_turns: 200,
+      stopped_at: null,
+      key_alias: "prod-alpha",
+      key_name: "sk-...alpha",
+    },
+  ],
   judged_count: 42,
   error_count: 1,
   judge_spend: 3.21,
@@ -110,19 +118,28 @@ const job = (overrides: Partial = {}): ShadowEvalJob => ({
         avg_judge_confidence: 0.8,
       },
     ],
+    by_key: [],
     overall_shadow_win_rate_pct: 48.0,
     overall_tie_rate_pct: 22.0,
   },
   created_at: "2026-08-07T00:00:00Z",
   ends_at: "2026-09-07T00:00:00Z",
-  stopped_at: null,
-  api_key_id: "hashed-key-abc",
-  key_alias: "prod-alpha",
-  key_name: "sk-...alpha",
   last_error: null,
   ...overrides,
 });
 
+const keyEntry = (
+  api_key_id: string,
+  overrides: Partial = {},
+): ShadowEvalJob["keys"][number] => ({
+  api_key_id,
+  max_turns: 200,
+  stopped_at: null,
+  key_alias: null,
+  key_name: null,
+  ...overrides,
+});
+
 const mockHooks = ({
   jobs = [],
   detailsById = {},
@@ -199,8 +216,8 @@ describe("ShadowEvalSection", () => {
   it("gives every active job its own card with a stop button, with the form still offered", () => {
     mockHooks({
       jobs: [
-        job({ job_id: "job-a", status: "running", api_key_id: "key-a" }),
-        job({ job_id: "job-b", status: "running", api_key_id: "key-b" }),
+        job({ job_id: "job-a", status: "running", keys: [keyEntry("key-a")] }),
+        job({ job_id: "job-b", status: "running", keys: [keyEntry("key-b")] }),
       ],
     });
     render();
@@ -342,7 +359,7 @@ describe("ShadowEvalSection", () => {
     expect(container).toBeEmptyDOMElement();
   });
 
-  it("keeps the start button disabled until key, router, and judge model are picked, then submits them", async () => {
+  it("keeps the start button disabled until key, router, and judge model are picked, then submits the key as a list", async () => {
     const user = userEvent.setup();
     const { start } = mockHooks({});
     render();
@@ -361,7 +378,7 @@ describe("ShadowEvalSection", () => {
     await user.click(screen.getByText("Start shadow eval"));
 
     const expectedBody = {
-      api_key_id: "hash-alpha",
+      api_key_ids: ["hash-alpha"],
       router_name: "gpt-auto",
       direction: "forward",
       shadow_percentage: 10,
@@ -396,7 +413,7 @@ describe("ShadowEvalSection", () => {
     await user.click(screen.getByText("Start shadow eval"));
 
     const expectedBody = {
-      api_key_id: "hash-alpha",
+      api_key_ids: ["hash-alpha"],
       router_name: "gpt-auto",
       direction: "reverse",
       baseline_model: "prod-claude",
@@ -429,9 +446,9 @@ describe("ShadowEvalSection", () => {
   });
 
   it("labels the shadowed key by alias, then masked name, then truncated hash", () => {
-    expect(shadowedKeyLabel(job())).toBe("prod-alpha");
-    expect(shadowedKeyLabel(job({ key_alias: null }))).toBe("sk-...alpha");
-    expect(shadowedKeyLabel(job({ key_alias: null, key_name: null }))).toBe("hashed-key…");
+    expect(shadowedKeyLabel(job().keys[0])).toBe("prod-alpha");
+    expect(shadowedKeyLabel(keyEntry("hashed-key-abc", { key_name: "sk-...alpha" }))).toBe("sk-...alpha");
+    expect(shadowedKeyLabel(keyEntry("hashed-key-abc"))).toBe("hashed-key…");
   });
 
   it("keeps an older job's verdicts reachable through the previous evaluations list", async () => {
diff --git a/ui/litellm-dashboard/src/app/(dashboard)/cost-optimization/_components/ShadowEvalSection.tsx b/ui/litellm-dashboard/src/app/(dashboard)/cost-optimization/_components/ShadowEvalSection.tsx
index 005636615b1..6d240a84c98 100644
--- a/ui/litellm-dashboard/src/app/(dashboard)/cost-optimization/_components/ShadowEvalSection.tsx
+++ b/ui/litellm-dashboard/src/app/(dashboard)/cost-optimization/_components/ShadowEvalSection.tsx
@@ -24,6 +24,7 @@ import {
   useStartShadowEval,
   useStopShadowEval,
   type ShadowEvalJob,
+  type ShadowEvalJobKey,
   type ShadowEvalSlice,
 } from "./useShadowEval";
 
@@ -50,19 +51,24 @@ const routerMatchedOrBeatPct = (
     ? 100 - results.overall_shadow_win_rate_pct
     : results.overall_shadow_win_rate_pct + results.overall_tie_rate_pct;
 
-export const shadowedKeyLabel = (job: ShadowEvalJob): string =>
-  job.key_alias || job.key_name || `${job.api_key_id.slice(0, 10)}…`;
+export const shadowedKeyLabel = (key: ShadowEvalJobKey): string =>
+  key.key_alias || key.key_name || `${key.api_key_id.slice(0, 10)}…`;
+
+const shadowedKeysLabel = (job: ShadowEvalJob): string =>
+  job.keys.length === 1 ? shadowedKeyLabel(job.keys[0]) : `${job.keys.length} keys`;
+
+const totalBudget = (job: ShadowEvalJob): number => job.keys.reduce((sum, key) => sum + key.max_turns, 0);
 
 const jobHeadline = (job: ShadowEvalJob): React.ReactNode =>
   job.direction === "reverse" ? (
     <>
       Comparing {job.router_name} to{" "}
       {job.baseline_model} on {job.shadow_percentage}% of{" "}
-      {shadowedKeyLabel(job)} traffic
+      {shadowedKeysLabel(job)} traffic
     
   ) : (
     <>
-      Shadowing {job.shadow_percentage}% of {shadowedKeyLabel(job)} traffic
+      Shadowing {job.shadow_percentage}% of {shadowedKeysLabel(job)} traffic
       via {job.router_name}
     
   );
@@ -178,7 +184,8 @@ const emptyResultsText = (job: ShadowEvalJob, resultsError: boolean): string =>
 
 const ResultsBody: React.FC<{ job: ShadowEvalJob; resultsError?: boolean }> = ({ job, resultsError = false }) => {
   const results = job.results;
-  if (!results || (results.by_tier.length === 0 && results.by_current_model.length === 0)) {
+  const stratifications = results ? [results.by_tier, results.by_current_model, results.by_key] : [];
+  if (!results || stratifications.every((slices) => slices.length === 0)) {
     return 
{emptyResultsText(job, resultsError)}
;
   }
   return (
@@ -224,7 +231,7 @@ const JobResults: React.FC<{
           

             
{jobHeadline(job)}

             

-              {(job.judged_count ?? 0).toLocaleString()} of {job.max_turns.toLocaleString()} turns judged ·{" "}
+              {(job.judged_count ?? 0).toLocaleString()} of {totalBudget(job).toLocaleString()} turns judged ·{" "}
               {(job.error_count ?? 0).toLocaleString()} errored · {usd(job.judge_spend ?? 0)} judge spend
               {active && remaining ? ` · ${remaining}` : ""}
             

@@ -386,14 +393,13 @@ const StartForm: React.FC = () => {
   const percentageValid = parsedPct >= 0.1 && parsedPct <= 100;
   const parsedMaxTurns = Number.parseInt(maxTurns, 10);
   const maxTurnsValid = parsedMaxTurns >= 1 && parsedMaxTurns <= 2000;
-  const filled =
-    [apiKeyId, routerName, judgeModel].every((field) => field !== "") &&
-    (direction === "forward" || baselineModel !== "");
+  const baselinePicked = direction === "forward" || baselineModel !== "";
+  const filled = [apiKeyId, routerName, judgeModel].every((field) => field !== "") && baselinePicked;
   const boundsValid = percentageValid && maxTurnsValid;
   const valid = Boolean(accessToken) && filled && boundsValid;
   const handleStart = () => {
     const startBody = {
-      api_key_id: apiKeyId,
+      api_key_ids: [apiKeyId],
       router_name: routerName,
       direction,
       ...(direction === "reverse" ? { baseline_model: baselineModel } : {}),
diff --git a/ui/litellm-dashboard/src/app/(dashboard)/cost-optimization/_components/useShadowEval.ts b/ui/litellm-dashboard/src/app/(dashboard)/cost-optimization/_components/useShadowEval.ts
index 7645fcc3346..eef98320e67 100644
--- a/ui/litellm-dashboard/src/app/(dashboard)/cost-optimization/_components/useShadowEval.ts
+++ b/ui/litellm-dashboard/src/app/(dashboard)/cost-optimization/_components/useShadowEval.ts
@@ -7,6 +7,7 @@ import { $api, fetchClient } from "@/lib/http/api";
 import type { components } from "@/lib/http/schema";
 
 export type ShadowEvalJob = components["schemas"]["ShadowEvalJobResponse"];
+export type ShadowEvalJobKey = components["schemas"]["ShadowEvalJobKeyResponse"];
 export type ShadowEvalSlice = components["schemas"]["ShadowEvalSlice"];
 export type StartShadowEvalRequest = components["schemas"]["StartShadowEvalRequest"];
 
diff --git a/ui/litellm-dashboard/src/lib/http/schema.d.ts b/ui/litellm-dashboard/src/lib/http/schema.d.ts
index 5794c60e97b..4c63586d4d6 100644
--- a/ui/litellm-dashboard/src/lib/http/schema.d.ts
+++ b/ui/litellm-dashboard/src/lib/http/schema.d.ts
@@ -816,7 +816,8 @@ export interface paths {
         };
         /**
          * List Shadow Eval Jobs
-         * @description List shadow eval jobs, newest first. Counts and results ride the detail endpoint only.
+         * @description List shadow eval jobs, newest first, each key with its attempt count so status is
+         *     accurate. Judged counts, spend, and results ride the detail endpoint only.
          */
         get: operations["list_shadow_eval_jobs_auto_router_shadow_eval_get"];
         put?: never;
@@ -838,20 +839,21 @@ export interface paths {
         put?: never;
         /**
          * Start Shadow Eval
-         * @description Start a shadow eval: duplicate a sampled slice of a key's live traffic against a second
-         *     arm, judge the two responses blind, and stratify win rates by tier and by the model that
-         *     served the real arm.
+         * @description Start a shadow eval: duplicate a sampled slice of one or more keys' live traffic against
+         *     a second arm, judge the two responses blind, and stratify win rates by tier, by the model
+         *     that served the real arm, and by key.
          *
-         *     A forward job answers whether the key should adopt router_name: it samples the requests
+         *     A forward job answers whether the keys should adopt router_name: it samples the requests
          *     the router did not serve and duplicates them through it. A reverse job answers whether a
          *     key already on the router still gains from it: it samples the requests the router did
          *     serve and duplicates them against baseline_model. A key can hold one active job per
          *     direction, so both questions can run at once.
          *
-         *     Shadow responses are never served to users. The job samples until it has judged
-         *     max_turns turns, reaches the end of its window, or is stopped; sampling changes
-         *     propagate to pods within about 10 seconds. Shadow and judge calls bill to the
-         *     shadowed key but are excluded from request counts and auto-router adoption metrics.
+         *     Shadow responses are never served to users. Each key samples until it has judged
+         *     max_turns turns of its own traffic, the job's window ends, or the job is stopped, so one
+         *     key running out of budget does not end sampling for the others; sampling changes
+         *     propagate to pods within about 10 seconds. Shadow and judge calls bill to the shadowed
+         *     key but are excluded from request counts and auto-router adoption metrics.
          */
         post: operations["start_shadow_eval_auto_router_shadow_eval_start_post"];
         delete?: never;
@@ -891,7 +893,12 @@ export interface paths {
         put?: never;
         /**
          * Stop Shadow Eval Job
-         * @description Stop an active shadow eval job. Attempts are kept; sampling halts within ~10s.
+         * @description Stop an active shadow eval job, every key it scopes at once. Attempts are kept;
+         *     sampling halts within ~10s. Keys that already stopped on their own budget keep the
+         *     stopped_at they earned. The statement is the whole state machine: it claims the job
+         *     only while a leg still samples inside the window with no stop recorded, so a racing
+         *     operator, a same-instant budget spend, and a repeat stop all read the same 400 with
+         *     the status the job actually holds.
          */
         post: operations["stop_shadow_eval_job_auto_router_shadow_eval__job_id__stop_post"];
         delete?: never;
@@ -33201,18 +33208,49 @@ export interface components {
             timeout?: number | null;
         };
         /**
-         * ShadowEvalJobResponse
-         * @description A shadow-eval job. Validates directly from the prisma record (job_id reads the
-         *     row's id); status is derived from stopped_at and ends_at, never stored, so no writer
-         *     anywhere can produce an inconsistent one. Aggregate fields are populated by the
-         *     detail endpoint only and stay None on list responses.
+         * ShadowEvalJobKeyResponse
+         * @description One key a job shadows, with its own budget and stop state.
          */
-        ShadowEvalJobResponse: {
+        ShadowEvalJobKeyResponse: {
             /**
              * Api Key Id
-             * @description The hashed virtual key whose traffic this job evaluates, and only that key's
+             * @description The hashed virtual key whose traffic this entry scopes
              */
             api_key_id: string;
+            /**
+             * Attempt Count
+             * @description This key's sampled attempts so far, judged and errored alike, the same count the sampler budgets against max_turns; populated on list and detail responses. Frozen at stopped_at once the key is stamped, so in-flight attempts landing after a stop never reclassify it
+             */
+            attempt_count?: number | null;
+            /**
+             * Key Alias
+             * @description Alias of the shadowed key, resolved from the key row at read time; None when unset or deleted
+             */
+            key_alias?: string | null;
+            /**
+             * Key Name
+             * @description Masked display name (sk-...) of the shadowed key, resolved at read time like key_alias
+             */
+            key_name?: string | null;
+            /**
+             * Max Turns
+             * @description This key's own sample budget, independent of its siblings'
+             */
+            max_turns: number;
+            /**
+             * Stopped At
+             * @description When this key's slot was stamped free, whether its own budget ran out, the window closed, or an operator stopped the job; status is derived, so a spent budget reads completed even while this is still unset
+             */
+            stopped_at?: string | null;
+        };
+        /**
+         * ShadowEvalJobResponse
+         * @description A shadow-eval job over one or more keys, each with its own budget and stop state;
+         *     status is derived from stopped_by, the keys' stop and budget state, and ends_at,
+         *     never stored, so no writer anywhere can produce an inconsistent one. Aggregate
+         *     fields are populated by the detail endpoint only and stay None on list responses.
+         */
+        ShadowEvalJobResponse: {
             /** Baseline Model */
             baseline_model?: string | null;
             /**
@@ -33251,22 +33289,15 @@ export interface components {
              */
             judged_count?: number | null;
             /**
-             * Key Alias
-             * @description Alias of the shadowed key, resolved from the key row at read time; None when unset or deleted
+             * Keys
+             * @description The keys whose traffic this job evaluates, and only those keys', each with its own budget
              */
-            key_alias?: string | null;
-            /**
-             * Key Name
-             * @description Masked display name (sk-...) of the shadowed key, resolved at read time like key_alias
-             */
-            key_name?: string | null;
+            keys: components["schemas"]["ShadowEvalJobKeyResponse"][];
             /**
              * Last Error
              * @description Most recent attempt error; detail endpoint only
              */
             last_error?: string | null;
-            /** Max Turns */
-            max_turns: number;
             /** @description Stratified verdicts; detail endpoint only */
             results?: components["schemas"]["ShadowEvalResult"] | null;
             /** Router Name */
@@ -33275,13 +33306,19 @@ export interface components {
             shadow_percentage: number;
             /**
              * Status
-             * @description A job whose window has passed reads completed even if a later sweep stamped
-             *     stopped_at; stopped means sampling ended before the window did.
+             * @description Three recorded facts, no history-guessing: a stop is stopped_by (the migration
+             *     backfills it for every job that displayed stopped when the column arrived, so the
+             *     pre-column population is closed), completion is the window passing or every key
+             *     spending its budget, and anything else is running. The all-keys-stamped fallback
+             *     covers only stops written by pre-column pods during a rolling deploy.
              * @enum {string}
              */
             readonly status: "running" | "completed" | "stopped";
-            /** Stopped At */
-            stopped_at?: string | null;
+            /**
+             * Stopped By
+             * @description The operator who stopped the job early, recorded by the stop endpoint; 'unknown' backfilled by migration for jobs that displayed stopped when the column arrived; None when the job ended on its own. Its presence is what makes a job read stopped rather than completed
+             */
+            stopped_by?: string | null;
         };
         /**
          * ShadowEvalResult
@@ -33290,9 +33327,14 @@ export interface components {
         ShadowEvalResult: {
             /**
              * By Current Model
-             * @description Sliced by the model that served the real arm: the key's incumbent models in forward mode, and in reverse the models the router itself picked
+             * @description Sliced by the model that served the real arm: the keys' incumbent models in forward mode, and in reverse the models the router itself picked
              */
             by_current_model: components["schemas"]["ShadowEvalSlice"][];
+            /**
+             * By Key
+             * @description One slice per scoped key that has judged verdicts, grouped on the raw key hash. Keys the job scopes but has not judged a turn for yet are absent rather than reported as zero
+             */
+            by_key: components["schemas"]["ShadowEvalSlice"][];
             /** By Tier */
             by_tier: components["schemas"]["ShadowEvalSlice"][];
             /** Overall Shadow Win Rate Pct */
@@ -33500,14 +33542,14 @@ export interface components {
         };
         /**
          * StartShadowEvalRequest
-         * @description Start duplicating a key's traffic for blind comparison against an auto-router.
+         * @description Start duplicating one or more keys' traffic for blind comparison against an auto-router.
          */
         StartShadowEvalRequest: {
             /**
-             * Api Key Id
-             * @description The hashed virtual key whose traffic will be shadowed. Shadow evaluation runs ONLY on this key's traffic; requests made with any other key are not sampled.
+             * Api Key Ids
+             * @description The hashed virtual keys whose traffic will be shadowed. Shadow evaluation runs ONLY on these keys' traffic; requests made with any other key are not sampled. Each key carries its own max_turns budget, so one key exhausting its budget leaves the others sampling. At most 100 keys per job, which also bounds every read the job's endpoints make.
              */
-            api_key_id: string;
+            api_key_ids: string[];
             /**
              * Baseline Model
              * @description Required when direction is reverse and rejected otherwise: the fixed model the router's own responses are judged against. Must be a plain model rather than another auto-router
@@ -33534,7 +33576,7 @@ export interface components {
             judge_model: string;
             /**
              * Max Turns
-             * @description Sample budget: the job judges at most this many turns, then completes. This is also the spend bound; expected judge cost is roughly max_turns times one judge call
+             * @description Per-key sample budget: the job judges at most this many turns of EACH scoped key's traffic, so a job over N keys judges at most N times max_turns turns. This is also the spend bound; expected judge cost is roughly that turn ceiling times one judge call
              * @default 200
              */
             max_turns: number;
@@ -37931,7 +37973,7 @@ export interface operations {
     list_shadow_eval_jobs_auto_router_shadow_eval_get: {
         parameters: {
             query?: {
-                /** @description Filter to jobs shadowing this key */
+                /** @description Filter to jobs that shadow this key, alone or alongside others */
                 api_key_id?: string | null;
                 /** @description Newest jobs to return */
                 limit?: number;

From 3d51eb378a027b17e4f123b3eb768b31150bcccb Mon Sep 17 00:00:00 2001
From: yuneng-jiang 
Date: Wed, 19 Aug 2026 14:07:50 -0700
Subject: [PATCH 9/9] refactor(ui): migrate the antd Alert call sites onto the
 shared Alert (#37513)

Moves all 33 antd Alert usages across 20 dashboard files onto
src/components/shared/Alert, following the composition the rest of the
dashboard already uses: message becomes AlertTitle, description becomes
AlertDescription, showIcon becomes a lucide icon child, and closable
becomes an AlertAction ghost button.

antd type="success" has no counterpart on the shared Alert, so the two
success sites land on the default variant with a CircleCheck icon, which
is what cloudzero_export_modal and CloudZeroIntegrationSettings already
do for the same case.

LoginPage's dismissible SSO notice moves into its own SsoEnabledNotice
component in the same file: antd's closable carried its own dismiss
state, and inlining it pushed LoginPageContent past the complexity
budget.

Four files lose their last antd symbol, so their no-restricted-imports
suppressions are pruned by hand. antd import sites drop from 115 to 111
across 107 to 103 files, and the no-restricted-imports ratchet drops
from 119 to 115 over 110 to 106 files.

One test asserted antd's own ant-alert-info class; it is repointed to
the shared Alert's text-info variant class, which keeps the same
"info, not warning" check. Every other colocated test passes untouched.
---
 ui/litellm-dashboard/eslint-suppressions.json |  18 --
 .../admin-panel/_components/AdminPanel.tsx    |  18 +-
 .../_components/MCPPermissionManagement.tsx   |  20 +-
 .../_components/UserEnvVarsModal.tsx          |  14 +-
 .../mcp-servers/_components/mcp_connect.tsx   |  50 ++--
 .../_components/mcp_server_edit.tsx           |  20 +-
 .../policies/_components/add_policy_form.tsx  |  60 ++---
 .../_components/policy_test_panel.tsx         |  10 +-
 .../_components/CreateVectorStore.tsx         |  38 +--
 .../_components/S3VectorsConfig.tsx           |  52 ++--
 .../_components/VectorStoreForm.tsx           | 246 +++++++++---------
 .../src/app/login/LoginPage.tsx               | 119 +++++----
 .../src/app/onboarding/OnboardingFormBody.tsx |  24 +-
 .../src/components/CreateUserButton.tsx       |  29 +--
 .../MCPSemanticFilterSettings.tsx             |  59 +++--
 .../src/components/add_model/AddModelForm.tsx |  18 +-
 .../src/components/add_pass_through.tsx       |  20 +-
 .../PassThroughGuardrailsSection.tsx          |  40 ++-
 .../user_search_modal.test.tsx                |   2 +-
 .../common_components/user_search_modal.tsx   |  18 +-
 .../organisms/RegenerateKeyModal.tsx          |  10 +-
 .../update_model_credentials_modal.tsx        |  18 +-
 22 files changed, 460 insertions(+), 443 deletions(-)

diff --git a/ui/litellm-dashboard/eslint-suppressions.json b/ui/litellm-dashboard/eslint-suppressions.json
index 98225de5d8b..09e602e269b 100644
--- a/ui/litellm-dashboard/eslint-suppressions.json
+++ b/ui/litellm-dashboard/eslint-suppressions.json
@@ -1404,11 +1404,6 @@
       "count": 2
     }
   },
-  "src/app/(dashboard)/vector-stores/_components/S3VectorsConfig.tsx": {
-    "no-restricted-imports": {
-      "count": 1
-    }
-  },
   "src/app/(dashboard)/vector-stores/_components/VectorStoreForm.tsx": {
     "no-nested-ternary": {
       "count": 2
@@ -1453,9 +1448,6 @@
     "local/no-complex-jsx-arrow": {
       "count": 1
     },
-    "no-restricted-imports": {
-      "count": 1
-    },
     "react-hooks/set-state-in-effect": {
       "count": 2
     }
@@ -1470,11 +1462,6 @@
       "count": 1
     }
   },
-  "src/app/onboarding/OnboardingFormBody.tsx": {
-    "no-restricted-imports": {
-      "count": 1
-    }
-  },
   "src/components/AIHub/ModelHubTable.test.tsx": {
     "max-params": {
       "count": 1
@@ -1998,11 +1985,6 @@
       "count": 1
     }
   },
-  "src/components/common_components/PassThroughGuardrailsSection.tsx": {
-    "no-restricted-imports": {
-      "count": 1
-    }
-  },
   "src/components/common_components/RateLimitTypeFormItem.test.tsx": {
     "no-restricted-imports": {
       "count": 1
diff --git a/ui/litellm-dashboard/src/app/(dashboard)/admin-panel/_components/AdminPanel.tsx b/ui/litellm-dashboard/src/app/(dashboard)/admin-panel/_components/AdminPanel.tsx
index 48eaa62e37b..e9d7190abf5 100644
--- a/ui/litellm-dashboard/src/app/(dashboard)/admin-panel/_components/AdminPanel.tsx
+++ b/ui/litellm-dashboard/src/app/(dashboard)/admin-panel/_components/AdminPanel.tsx
@@ -7,8 +7,8 @@ import { Alert, AlertDescription, AlertTitle } from "@/components/shared/Alert";
 import { Button } from "@/components/ui/button";
 import { Card } from "@/components/ui/card";
 import { Table, TableBody, TableCell, TableHead, TableHeader, TableRow } from "@/components/ui/table";
-import { Alert as AntdAlert, Modal, Space, Tabs, Typography } from "antd";
-import { Info } from "lucide-react";
+import { Modal, Space, Tabs, Typography } from "antd";
+import { Info, TriangleAlert } from "lucide-react";
 import React, { useEffect, useState } from "react";
 import NewBadge from "@/components/common_components/NewBadge";
 import { useBaseUrl } from "@/components/constants";
@@ -223,12 +223,14 @@ const AdminPanel: React.FC = ({ proxySettings }) => {
         <>
           
              ✨ Security Settings
-            
+            
+              
+              SSO Configuration Deprecated
+              
+                Editing SSO Settings on this page is deprecated and will be removed in a future version. Please use the
+                SSO Settings tab for SSO configuration.
+              
+            
             
 = ({
           )}
 
           {showInternalDelegatePkceWarning && (
-            
+            
+              
+              Internal server with upstream OAuth delegation
+              
+                This MCP server is configured as internal-only but delegates auth to upstream. Anonymous users will be
+                able to reach the upstream OAuth2 /authorize flow without a LiteLLM session. Ensure your upstream
+                provider and network enforce access controls.
+              
+            
           )}
 
            = ({ server, open, acces
             
           

         ) : isError ? (
-          
+          
+            
+            Failed to load env vars
+          
         ) : required.length === 0 ? (
-          
+          
+            
+            No per-user fields configured for this server.
+          
         ) : (
           <>
             
diff --git a/ui/litellm-dashboard/src/app/(dashboard)/mcp-servers/_components/mcp_connect.tsx b/ui/litellm-dashboard/src/app/(dashboard)/mcp-servers/_components/mcp_connect.tsx
index bb739901faa..8aeb94404f2 100644
--- a/ui/litellm-dashboard/src/app/(dashboard)/mcp-servers/_components/mcp_connect.tsx
+++ b/ui/litellm-dashboard/src/app/(dashboard)/mcp-servers/_components/mcp_connect.tsx
@@ -1,10 +1,22 @@
 /* eslint-disable react/no-unescaped-entities */
 
 import React, { useState } from "react";
-import { Card, Typography, Space, Alert, Switch } from "antd";
+import { Card, Typography, Space, Switch } from "antd";
+import { Alert, AlertDescription, AlertTitle } from "@/components/shared/Alert";
 import { Button } from "@/components/ui/button";
 import { Tabs, TabsContent, TabsList, TabsTrigger } from "@/components/ui/tabs";
-import { CopyIcon, Code, Terminal, Globe, CheckIcon, ExternalLinkIcon, KeyIcon, ServerIcon, Zap } from "lucide-react";
+import {
+  CopyIcon,
+  Code,
+  Terminal,
+  Globe,
+  CheckIcon,
+  ExternalLinkIcon,
+  Info,
+  KeyIcon,
+  ServerIcon,
+  Zap,
+} from "lucide-react";
 import { getProxyBaseUrl } from "@/components/networking";
 import { copyToClipboard as utilCopyToClipboard } from "@/utils/dataUtils";
 
@@ -69,25 +81,21 @@ const FeatureCard: React.FC = ({
             
           

           {useServerHeader && (
-            
-                  

-                    Option 1: Get a specific server: "{serverName.replace(/\s+/g, "_")}"
-                  

-                  

-                    Option 2: Get a group of MCPs: "dev-group"
-                  

-                  

-                    You can also mix both: "Server1,dev-group"
-                  

-                
-              }
-            />
+            
+              
+              Two Options
+              
+                

+                  Option 1: Get a specific server: "{serverName.replace(/\s+/g, "_")}"
+                

+                

+                  Option 2: Get a group of MCPs: "dev-group"
+                

+                

+                  You can also mix both: "Server1,dev-group"
+                

+              
+            
           )}
         
       )}
diff --git a/ui/litellm-dashboard/src/app/(dashboard)/mcp-servers/_components/mcp_server_edit.tsx b/ui/litellm-dashboard/src/app/(dashboard)/mcp-servers/_components/mcp_server_edit.tsx
index eb462f788a5..94b66f1bdfa 100644
--- a/ui/litellm-dashboard/src/app/(dashboard)/mcp-servers/_components/mcp_server_edit.tsx
+++ b/ui/litellm-dashboard/src/app/(dashboard)/mcp-servers/_components/mcp_server_edit.tsx
@@ -1,5 +1,7 @@
 import React, { useState, useEffect } from "react";
-import { Select, Tooltip, Input, InputNumber, Alert } from "antd";
+import { Select, Tooltip, Input, InputNumber } from "antd";
+import { TriangleAlert } from "lucide-react";
+import { Alert, AlertDescription, AlertTitle } from "@/components/shared/Alert";
 import { FormProvider, useForm } from "react-hook-form";
 import { InfoCircleOutlined } from "@ant-design/icons";
 import { Button } from "@/components/ui/button";
@@ -1041,13 +1043,15 @@ const MCPServerEdit: React.FC = ({
                 {!isStdioTransport && isOAuthAuthType && (
                   <>
                     {!oauthFlowTypeValue && !isDelegateAuth && (
-                      
+                      
+                        
+                        This server has no OAuth flow set
+                        
+                          Choose Machine-to-Machine (M2M) or Interactive (PKCE) so LiteLLM authenticates it the way you
+                          intend, then save. Until it is set, LiteLLM falls back to interactive per-user auth and treats
+                          a machine-to-machine credential shape conservatively.
+                        
+                      
                     )}
                      = ({
 
         {selectedMode === "flow_builder" && (
           
+          >
+            
+              You'll be redirected to the full-screen Flow Builder to design your policy logic visually.
+            
+          
         )}
 
         

@@ -457,35 +461,33 @@ const AddPolicyForm: React.FC = ({
             
 
             {resolvedGuardrails.length > 0 && (
-              
-                    
-                      These are the final guardrails that will be applied (including inheritance):
-                    
-                    

-                      {resolvedGuardrails.map((g) => (
-                        
-                          {g}
-                        
-                      ))}
-                    

+              
+                
+                Resolved Guardrails
+                
+                  
+                    These are the final guardrails that will be applied (including inheritance):
+                  
+                  

+                    {resolvedGuardrails.map((g) => (
+                      
+                        {g}
+                      
+                    ))}
                   

-                }
-                type="info"
-                showIcon
-              />
+                
+              
             )}
 
             
 
-            
+            
+              
+              Model Scope
+              
+                By default, this policy will run on all models. You can optionally restrict it to specific models below.
+              
+            
 
             

               Model Condition Type
diff --git a/ui/litellm-dashboard/src/app/(dashboard)/policies/_components/policy_test_panel.tsx b/ui/litellm-dashboard/src/app/(dashboard)/policies/_components/policy_test_panel.tsx
index 90642d23a46..9c234a7a70c 100644
--- a/ui/litellm-dashboard/src/app/(dashboard)/policies/_components/policy_test_panel.tsx
+++ b/ui/litellm-dashboard/src/app/(dashboard)/policies/_components/policy_test_panel.tsx
@@ -1,6 +1,8 @@
 import React, { useState, useEffect } from "react";
 import { useForm } from "react-hook-form";
-import { Alert, Empty } from "antd";
+import { Empty } from "antd";
+import { CircleAlert } from "lucide-react";
+import { Alert, AlertDescription, AlertTitle } from "@/components/shared/Alert";
 import { resolvePoliciesCall, teamListCall, keyListCall, modelAvailableCall } from "@/components/networking";
 import useAuthorized from "@/app/(dashboard)/hooks/useAuthorized";
 import { FieldGroup } from "@/components/shared/form/field";
@@ -325,7 +327,11 @@ const PolicyTestPanel: React.FC = ({ accessToken }) => {
       )}
 
       {hasSearched && !result && !isLoading && (
-        
+        
+          
+          Error
+          Failed to resolve policies. Check the proxy logs.
+        
       )}
     

   );
diff --git a/ui/litellm-dashboard/src/app/(dashboard)/vector-stores/_components/CreateVectorStore.tsx b/ui/litellm-dashboard/src/app/(dashboard)/vector-stores/_components/CreateVectorStore.tsx
index 93aaa90ef7c..271ef31a68e 100644
--- a/ui/litellm-dashboard/src/app/(dashboard)/vector-stores/_components/CreateVectorStore.tsx
+++ b/ui/litellm-dashboard/src/app/(dashboard)/vector-stores/_components/CreateVectorStore.tsx
@@ -1,9 +1,10 @@
 import React, { useState } from "react";
-import { Upload, Alert } from "antd";
+import { Upload } from "antd";
 import { toast } from "@/lib/toast";
 import { InboxOutlined } from "@ant-design/icons";
 import type { UploadProps } from "antd";
-import { CircleHelp } from "lucide-react";
+import { CircleCheck, CircleHelp, X } from "lucide-react";
+import { Alert, AlertAction, AlertDescription, AlertTitle } from "@/components/shared/Alert";
 import { ragIngestCall } from "@/components/networking";
 import { DocumentUpload, RAGIngestResponse } from "@/components/vector_store_management/types";
 import DocumentsTable from "./DocumentsTable";
@@ -359,22 +360,23 @@ const CreateVectorStore: React.FC = ({ accessToken, onSu
 
         {/* Success Message */}
         {ingestResults.length > 0 && (
-          
-                

-                  Vector Store ID: {ingestResults[0]?.vector_store_id}
-                

-                

-                  Documents Ingested: {ingestResults.length}
-                

-              

-            }
-            type="success"
-            showIcon
-            closable
-          />
+          
+            
+            Vector Store Created Successfully
+            
+              

+                Vector Store ID: {ingestResults[0]?.vector_store_id}
+              

+              

+                Documents Ingested: {ingestResults.length}
+              

+            
+            
+              
+            
+          
         )}
       
     
diff --git a/ui/litellm-dashboard/src/app/(dashboard)/vector-stores/_components/S3VectorsConfig.tsx b/ui/litellm-dashboard/src/app/(dashboard)/vector-stores/_components/S3VectorsConfig.tsx
index 3e286ee9b92..cdc63e8b7ec 100644
--- a/ui/litellm-dashboard/src/app/(dashboard)/vector-stores/_components/S3VectorsConfig.tsx
+++ b/ui/litellm-dashboard/src/app/(dashboard)/vector-stores/_components/S3VectorsConfig.tsx
@@ -1,6 +1,6 @@
 import React, { useState, useEffect } from "react";
-import { Alert } from "antd";
-import { CircleHelp } from "lucide-react";
+import { CircleHelp, Info } from "lucide-react";
+import { Alert, AlertDescription, AlertTitle } from "@/components/shared/Alert";
 import { fetchAvailableModels, ModelGroup } from "@/components/llm_calls/fetch_models";
 import { Field, FieldError, FieldLabel } from "@/components/shared/form/field";
 import {
@@ -72,32 +72,28 @@ const S3VectorsConfig: React.FC = ({ accessToken, provider
 
   return (
     
-      
-            
AWS S3 Vectors allows you to store and query vector embeddings directly in S3:

-            
-          
-        }
-        type="info"
-        showIcon
-        style={{ marginBottom: "16px" }}
-      />
+      
+        
+        AWS S3 Vectors Setup
+        
+          
AWS S3 Vectors allows you to store and query vector embeddings directly in S3:

+          
    
+            
  • Vector buckets and indexes will be automatically created if they don't exist
    • 
+            
  • Vector dimensions are auto-detected from your selected embedding model
    • 
+            
  • Ensure your AWS credentials have permissions for S3 Vectors operations
    • 
+            
  • 
+              Learn more:{" "}
+              
+                AWS S3 Vectors Documentation
+              
+            
    • 
+          

+        
+      
 
       
         
diff --git a/ui/litellm-dashboard/src/app/(dashboard)/vector-stores/_components/VectorStoreForm.tsx b/ui/litellm-dashboard/src/app/(dashboard)/vector-stores/_components/VectorStoreForm.tsx
index 05c829d79df..ff6c10bf428 100644
--- a/ui/litellm-dashboard/src/app/(dashboard)/vector-stores/_components/VectorStoreForm.tsx
+++ b/ui/litellm-dashboard/src/app/(dashboard)/vector-stores/_components/VectorStoreForm.tsx
@@ -1,6 +1,7 @@
 import React, { useState, useEffect } from "react";
-import { Modal, Alert } from "antd";
-import { CircleHelp, Eye, EyeOff } from "lucide-react";
+import { Modal } from "antd";
+import { CircleHelp, Eye, EyeOff, Info } from "lucide-react";
+import { Alert, AlertDescription, AlertTitle } from "@/components/shared/Alert";
 import { useWatch } from "react-hook-form";
 import { z } from "zod/v4";
 import { CredentialItem, vectorStoreCreateCall } from "@/components/networking";
@@ -310,142 +311,129 @@ const VectorStoreForm: React.FC = ({
             
 
             {selectedProvider === "pg_vector" && (
-              
-                    
LiteLLM provides a server to connect to PG Vector. To use this provider:

-                    
    
-                      
  1. 
-                        Deploy the litellm-pgvector server from:{" "}
-                        
-                          https://github.com/BerriAI/litellm-pgvector
-                        
-                      
    2. 
-                      
  2. Configure your PostgreSQL database with pgvector extension
    3. 
-                      
  3. Start the server and note the API base URL and API key
    4. 
-                      
  4. Enter those details in the fields below
    5. 
-                    

-                  
-                }
-                type="info"
-                showIcon
-              />
+              
+                
+                PG Vector Setup Required
+                
+                  
LiteLLM provides a server to connect to PG Vector. To use this provider:

+                  
    
+                    
  1. 
+                      Deploy the litellm-pgvector server from:{" "}
+                      
+                        https://github.com/BerriAI/litellm-pgvector
+                      
+                    
    2. 
+                    
  2. Configure your PostgreSQL database with pgvector extension
    3. 
+                    
  3. Start the server and note the API base URL and API key
    4. 
+                    
  4. Enter those details in the fields below
    5. 
+                  

+                
+              
             )}
 
             {selectedProvider === "valkey" && (
-              
-                    

-                      LiteLLM searches documents you have already stored in Valkey. It does not create the index or
-                      upload documents for you. Before creating this vector store, make sure:
-                    

-                    
    
-                      
  1. 
-                        Your Valkey server has vector search enabled (the valkey-search module, included in the
-                        valkey-bundle image and in AWS ElastiCache / MemoryDB for Valkey)
-                      
    2. 
-                      
  2. 
-                        You have already created a search index and loaded your documents and their embeddings into it.
-                        Enter that index name as the Vector Store ID
-                      
    3. 
-                      
  3. 
-                        You know which embedding model created those stored embeddings. That model must be added to this
-                        proxy under Models so you can pick it below. Using a different model returns wrong results
-                      
    4. 
-                      
  4. 
-                        You know the field names your documents use for their text and their embedding. If they are not
-                        "text" and "embedding", set them below
-                      
    5. 
-                    

-                    

-                      When a query comes in, LiteLLM converts it to an embedding with the model below and returns the
-                      closest matching documents from your index.
-                    

-                  
-                }
-                type="info"
-                showIcon
-              />
+              
+                
+                Valkey Setup Required
+                
+                  

+                    LiteLLM searches documents you have already stored in Valkey. It does not create the index or upload
+                    documents for you. Before creating this vector store, make sure:
+                  

+                  
    
+                    
  1. 
+                      Your Valkey server has vector search enabled (the valkey-search module, included in the
+                      valkey-bundle image and in AWS ElastiCache / MemoryDB for Valkey)
+                    
    2. 
+                    
  2. 
+                      You have already created a search index and loaded your documents and their embeddings into it.
+                      Enter that index name as the Vector Store ID
+                    
    3. 
+                    
  3. 
+                      You know which embedding model created those stored embeddings. That model must be added to this
+                      proxy under Models so you can pick it below. Using a different model returns wrong results
+                    
    4. 
+                    
  4. 
+                      You know the field names your documents use for their text and their embedding. If they are not
+                      "text" and "embedding", set them below
+                    
    5. 
+                  

+                  

+                    When a query comes in, LiteLLM converts it to an embedding with the model below and returns the
+                    closest matching documents from your index.
+                  

+                
+              
             )}
 
             {selectedProvider === "vertex_rag_engine" && (
-              
-                    
To use Vertex AI RAG Engine:

-                    

-                      Note: Google Cloud has renamed this to "RAG Engine" in its console — the steps below
-                      still apply.
-                    

-                    
    
-                      
  1. 
-                        Set up your Vertex AI RAG Engine corpus following the guide:{" "}
-                        
-                          Vertex AI RAG Engine Overview
-                        
-                      
    2. 
-                      
  2. Create a corpus in your Google Cloud project
    3. 
-                      
  3. 
-                        Note the corpus ID from the Vertex AI console (now labeled "RAG Engine" in Google
-                        Cloud)
-                      
    4. 
-                      
  4. Enter the corpus ID in the Vector Store ID field below
    5. 
-                    

-                  
-                }
-                type="info"
-                showIcon
-              />
+              
+                
+                Vertex AI RAG Engine Setup
+                
+                  
To use Vertex AI RAG Engine:

+                  

+                    Note: Google Cloud has renamed this to "RAG Engine" in its console — the steps below still
+                    apply.
+                  

+                  
    
+                    
  1. 
+                      Set up your Vertex AI RAG Engine corpus following the guide:{" "}
+                      
+                        Vertex AI RAG Engine Overview
+                      
+                    
    2. 
+                    
  2. Create a corpus in your Google Cloud project
    3. 
+                    
  3. 
+                      Note the corpus ID from the Vertex AI console (now labeled "RAG Engine" in Google Cloud)
+                    
    4. 
+                    
  4. Enter the corpus ID in the Vector Store ID field below
    5. 
+                  

+                
+              
             )}
 
             {selectedProvider === "vertex_ai/search_api" && (
-              
-                    
To use Vertex AI Search (Discovery Engine):

-                    

-                      Note: Google Cloud has renamed this to "Agent Search" in its console — the steps below
-                      still apply.
-                    

-                    
    
-                      
  1. 
-                        Enable the Discovery Engine API on your Google Cloud project and create a data store following
-                        the guide:{" "}
-                        
-                          Create a Vertex AI Search data store
-                        
-                      
    2. 
-                      
  2. Pick a supported location: global, us, or eu
    3. 
-                      
  3. 
-                        For most data store types (Cloud Storage, BigQuery, Media): copy the data store ID and enter it
-                        in the Vector Store ID field below.
-                      
    4. 
-                      
  4. 
-                        For website, healthcare, and connector-based sources (Drive, Gmail, Slack, Jira, etc.): create a
-                        search app on top of the data store, then copy the Engine ID and enter it in
-                        the Engine ID field. The Vector Store ID is still required as the LiteLLM-side name for this
-                        record, but it isn't used in the GCP URL when Engine ID is set.
-                      
    5. 
-                    

-                  
-                }
-                type="info"
-                showIcon
-              />
+              
+                
+                Vertex AI Search Setup
+                
+                  
To use Vertex AI Search (Discovery Engine):

+                  

+                    Note: Google Cloud has renamed this to "Agent Search" in its console — the steps below
+                    still apply.
+                  

+                  
    
+                    
  1. 
+                      Enable the Discovery Engine API on your Google Cloud project and create a data store following the
+                      guide:{" "}
+                      
+                        Create a Vertex AI Search data store
+                      
+                    
    2. 
+                    
  2. Pick a supported location: global, us, or eu
    3. 
+                    
  3. 
+                      For most data store types (Cloud Storage, BigQuery, Media): copy the data store ID and enter it in
+                      the Vector Store ID field below.
+                    
    4. 
+                    
  4. 
+                      For website, healthcare, and connector-based sources (Drive, Gmail, Slack, Jira, etc.): create a
+                      search app on top of the data store, then copy the Engine ID and enter it in the
+                      Engine ID field. The Vector Store ID is still required as the LiteLLM-side name for this record,
+                      but it isn't used in the GCP URL when Engine ID is set.
+                    
    5. 
+                  

+                
+              
             )}
 
             ;
 
+function SsoEnabledNotice() {
+  const [isDismissed, setIsDismissed] = useState(false);
+  if (isDismissed) return null;
+
+  return (
+    
+      
+      
+        Single Sign-On (SSO) is enabled. LiteLLM no longer automatically redirects to the SSO login flow upon loading
+        this page. To re-enable auto-redirect-to-SSO, set{" "}
+        AUTO_REDIRECT_UI_LOGIN_TO_SSO=true in your
+        environment configuration.
+      
+      
+        
+      
+    
+  );
+}
+
 function LoginPageContent() {
   const [isLoading, setIsLoading] = useState(true);
   const { data: uiConfig, isLoading: isConfigLoading } = useUIConfig();
@@ -170,22 +192,19 @@ function LoginPageContent() {
                 
🚅 LiteLLM

               
 
-              
-                    

-                      The Admin UI has been disabled by the administrator. To re-enable it, please update the following
-                      environment variable:
-                    

-                    

-                      DISABLE_ADMIN_UI=False
-                    

-                  
-                }
-                type="warning"
-                showIcon
-              />
+              
+                
+                Admin UI Disabled
+                
+                  

+                    The Admin UI has been disabled by the administrator. To re-enable it, please update the following
+                    environment variable:
+                  

+                  

+                    DISABLE_ADMIN_UI=False
+                  

+                
+              
             
           
         
@@ -209,31 +228,32 @@ function LoginPageContent() {
               
 
               {!uiConfig?.hide_default_credentials_hint && (
-                
-                      

-                        By default, Username is admin{" "}
-                        and Password is your set LiteLLM Proxy
-                        MASTER_KEY.
-                      

-                      

-                        Need to set UI credentials or SSO?{" "}
-                        
-                          Check the documentation
-                        
-                        .
-                      

-                    
-                  }
-                  type="info"
-                  icon={}
-                  showIcon
-                />
+                
+                  
+                  Default Credentials
+                  
+                    

+                      By default, Username is admin and
+                      Password is your set LiteLLM Proxy
+                      MASTER_KEY.
+                    

+                    

+                      Need to set UI credentials or SSO?{" "}
+                      
+                        Check the documentation
+                      
+                      .
+                    

+                  
+                
               )}
 
-              {error && }
+              {error && (
+                
+                  
+                  {error}
+                
+              )}
 
               

                 
@@ -326,22 +346,7 @@ function LoginPageContent() {
                 
               
             
-            {uiConfig?.sso_configured && (
-              
-                    Single Sign-On (SSO) is enabled. LiteLLM no longer automatically redirects to the SSO login flow
-                    upon loading this page. To re-enable auto-redirect-to-SSO, set{" "}
-                    AUTO_REDIRECT_UI_LOGIN_TO_SSO=true{" "}
-                    in your environment configuration.
-                  
-                }
-              />
-            )}
+            {uiConfig?.sso_configured && }
           
         
       
diff --git a/ui/litellm-dashboard/src/app/onboarding/OnboardingFormBody.tsx b/ui/litellm-dashboard/src/app/onboarding/OnboardingFormBody.tsx
index 1f2441c4870..0a03cc6a591 100644
--- a/ui/litellm-dashboard/src/app/onboarding/OnboardingFormBody.tsx
+++ b/ui/litellm-dashboard/src/app/onboarding/OnboardingFormBody.tsx
@@ -1,6 +1,7 @@
-import { Alert } from "antd";
+import { CircleAlert, Info } from "lucide-react";
 import React from "react";
 import { z } from "zod/v4";
+import { Alert, AlertDescription, AlertTitle } from "@/components/shared/Alert";
 import { PasswordInput } from "@/components/shared/PasswordInput";
 import { Field, FieldLabel, FieldGroup } from "@/components/shared/form/field";
 import { FormField } from "@/components/shared/form/FormField";
@@ -46,11 +47,10 @@ export function OnboardingFormBody({ variant, userEmail, isPending, claimError,
           

 
           {variant === "signup" && (
-            
+              
+              SSO
+              
                 

                   SSO is under the Enterprise Tier.
                   
                 

-              }
-              showIcon
-            />
+              
+            
           )}
 
           

@@ -84,7 +83,12 @@ export function OnboardingFormBody({ variant, userEmail, isPending, claimError,
               
             
 
-            {claimError && }
+            {claimError && (
+              
+                
+                {claimError}
+              
+            )}
 
             

               
+              
+            
           )}
 
           {updateError && (
-            
+            
+              Could not update settings
+              {updateError instanceof Error && {updateError.message}}
+            
           )}
 
           
diff --git a/ui/litellm-dashboard/src/components/add_model/AddModelForm.tsx b/ui/litellm-dashboard/src/components/add_model/AddModelForm.tsx
index 4cb60dc116a..c0c73fbb762 100644
--- a/ui/litellm-dashboard/src/components/add_model/AddModelForm.tsx
+++ b/ui/litellm-dashboard/src/components/add_model/AddModelForm.tsx
@@ -5,7 +5,9 @@ import { all_admin_roles, isUserTeamAdminForAnyTeam } from "@/utils/roles";
 import { modelCreationScope } from "@/utils/modelPermissions";
 import { Switch } from "@/components/ui/switch";
 import { Field, FieldLabel } from "@/components/shared/form/field";
-import { Select as AntdSelect, Card, Col, Modal, Row, Tooltip, Typography, Alert } from "antd";
+import { Select as AntdSelect, Card, Col, Modal, Row, Tooltip, Typography } from "antd";
+import { Info } from "lucide-react";
+import { Alert, AlertDescription, AlertTitle } from "@/components/shared/Alert";
 import { Button } from "@/components/ui/button";
 import type { UploadProps } from "antd/es/upload";
 import React, { useEffect, useMemo, useState } from "react";
@@ -168,13 +170,13 @@ const AddModelForm: React.FC = ({
                       )}
                     
                     {!teamAdminSelectedTeam && (
-                      
+                      
+                        
+                        Team Selection Required
+                        
+                          As a team admin, you need to select your team first before adding models.
+                        
+                      
                     )}
                   
                 )}
diff --git a/ui/litellm-dashboard/src/components/add_pass_through.tsx b/ui/litellm-dashboard/src/components/add_pass_through.tsx
index 9f95e7f27ac..09ddac2cf42 100644
--- a/ui/litellm-dashboard/src/components/add_pass_through.tsx
+++ b/ui/litellm-dashboard/src/components/add_pass_through.tsx
@@ -1,8 +1,9 @@
 "use client";
 
 import React, { useState } from "react";
-import { Modal, Alert } from "antd";
-import { CircleHelp, Plug } from "lucide-react";
+import { Modal } from "antd";
+import { CircleHelp, Info, Plug } from "lucide-react";
+import { Alert, AlertDescription, AlertTitle } from "@/components/shared/Alert";
 import { useWatch } from "react-hook-form";
 import { z } from "zod/v4";
 
@@ -156,13 +157,14 @@ const AddPassThroughEndpoint: React.FC = ({
           }}
         >
           

-            
+            
+              
+              What is a Pass-Through Endpoint?
+              
+                Route requests from your LiteLLM proxy to any external API. Perfect for custom models, image generation
+                APIs, or any service you want to proxy through LiteLLM.
+              
+            
 
             
               
diff --git a/ui/litellm-dashboard/src/components/common_components/PassThroughGuardrailsSection.tsx b/ui/litellm-dashboard/src/components/common_components/PassThroughGuardrailsSection.tsx
index f51629b7ee1..a544a8be32b 100644
--- a/ui/litellm-dashboard/src/components/common_components/PassThroughGuardrailsSection.tsx
+++ b/ui/litellm-dashboard/src/components/common_components/PassThroughGuardrailsSection.tsx
@@ -1,6 +1,6 @@
 import React from "react";
-import { Alert } from "antd";
-import { CircleHelp } from "lucide-react";
+import { CircleHelp, Info } from "lucide-react";
+import { Alert, AlertDescription, AlertTitle } from "@/components/shared/Alert";
 
 import GuardrailSelector from "../guardrails/GuardrailSelector";
 import { TagsInput } from "@/app/(dashboard)/guardrails/_components/content_filter/TagsInput";
@@ -67,21 +67,20 @@ const PassThroughGuardrailsSection: React.FC
           endpoints.
         

 
-        
-              Field-Level Targeting{" "}
-              
-                (Learn More)
-              
-            
-          }
-          description={
+        
+          
+          
+            Field-Level Targeting{" "}
+            
+              (Learn More)
+            
+          
+          
             

               

                 Optionally specify which fields to check. If left empty, the entire request/response is sent to the
@@ -100,11 +99,8 @@ const PassThroughGuardrailsSection: React.FC
                 

               

             

-          }
-          type="info"
-          showIcon
-          className="mb-4"
-        />
+          
+        
 
         
           
diff --git a/ui/litellm-dashboard/src/components/common_components/user_search_modal.test.tsx b/ui/litellm-dashboard/src/components/common_components/user_search_modal.test.tsx
index cbaf00b7131..946bbdd649b 100644
--- a/ui/litellm-dashboard/src/components/common_components/user_search_modal.test.tsx
+++ b/ui/litellm-dashboard/src/components/common_components/user_search_modal.test.tsx
@@ -74,7 +74,7 @@ describe("UserSearchModal", () => {
     expect(notice).toHaveTextContent(/users that already exist/i);
     expect(notice).toHaveTextContent(/ask a proxy admin to create their account first/i);
     // info, not warning: a warning here would read as an error state on an empty form
-    expect(notice.className).toMatch(/ant-alert-info/);
+    expect(notice.className).toMatch(/text-info/);
   });
 });
 
diff --git a/ui/litellm-dashboard/src/components/common_components/user_search_modal.tsx b/ui/litellm-dashboard/src/components/common_components/user_search_modal.tsx
index aeead9d82e6..e089b15addc 100644
--- a/ui/litellm-dashboard/src/components/common_components/user_search_modal.tsx
+++ b/ui/litellm-dashboard/src/components/common_components/user_search_modal.tsx
@@ -1,5 +1,7 @@
 import { useState } from "react";
-import { Modal, Alert } from "antd";
+import { Modal } from "antd";
+import { Info } from "lucide-react";
+import { Alert, AlertTitle } from "@/components/shared/Alert";
 import { UserAddOutlined } from "@ant-design/icons";
 import { useDebouncedCallback } from "@tanstack/react-pacer/debouncer";
 import { useForm } from "react-hook-form";
@@ -202,13 +204,13 @@ const UserSearchModal: React.FC = ({
     
       
         
-          
+          
+            
+            
+              Search selects from users that already exist. To add someone new, ask a proxy admin to create their
+              account first.
+            
+          
 
           
             
diff --git a/ui/litellm-dashboard/src/components/organisms/RegenerateKeyModal.tsx b/ui/litellm-dashboard/src/components/organisms/RegenerateKeyModal.tsx
index d717c11de86..88554ddba89 100644
--- a/ui/litellm-dashboard/src/components/organisms/RegenerateKeyModal.tsx
+++ b/ui/litellm-dashboard/src/components/organisms/RegenerateKeyModal.tsx
@@ -1,8 +1,9 @@
 import useAuthorized from "@/app/(dashboard)/hooks/useAuthorized";
 import { CheckOutlined, CopyOutlined, SyncOutlined } from "@ant-design/icons";
-import { Alert, Modal, Space } from "antd";
+import { Alert, AlertTitle } from "@/components/shared/Alert";
+import { Modal, Space } from "antd";
 import { Button } from "@/components/ui/button";
-import { CircleHelp } from "lucide-react";
+import { CircleHelp, TriangleAlert } from "lucide-react";
 import React, { useEffect, useMemo, useState } from "react";
 import { useWatch } from "react-hook-form";
 import { CopyToClipboard } from "react-copy-to-clipboard";
@@ -184,7 +185,10 @@ export function RegenerateKeyModal({ selectedToken, visible, onClose, onKeyUpdat
     >
       {regeneratedKey ? (
         

-          
+          
+            
+            Save it now, you will not see it again
+          
 
           

             Key Alias
diff --git a/ui/litellm-dashboard/src/components/update_model_credentials_modal.tsx b/ui/litellm-dashboard/src/components/update_model_credentials_modal.tsx
index b0a19e7ea05..a436ba55c01 100644
--- a/ui/litellm-dashboard/src/components/update_model_credentials_modal.tsx
+++ b/ui/litellm-dashboard/src/components/update_model_credentials_modal.tsx
@@ -1,10 +1,12 @@
-import { Alert, Modal, Typography } from "antd";
+import { Modal, Typography } from "antd";
+import { TriangleAlert } from "lucide-react";
 import { useState } from "react";
 import { z } from "zod/v4";
 import { modelPatchUpdateCall } from "./networking";
 import { toast } from "@/lib/toast";
 import { FieldGroup } from "@/components/shared/form/field";
 import { FormField } from "@/components/shared/form/FormField";
+import { Alert, AlertTitle } from "@/components/shared/Alert";
 import { PasswordInput } from "@/components/shared/PasswordInput";
 import { Button } from "@/components/ui/button";
 import { UiLoadingSpinner } from "@/components/ui/ui-loading-spinner";
@@ -74,12 +76,14 @@ export default function UpdateModelCredentialsModal({
         Update this model's API key. Only the new key is sent; the rest of the deployment configuration is left
         untouched.
       
-      
+      
+        
+        
+          Only the API key is rotated here. Models that authenticate with an Azure AD token, AWS credentials, or a
+          Vertex service-account JSON aren't supported yet; update those from the model's LiteLLM Params for
+          now.
+        
+