diff --git a/.github/scripts/close_duplicate_issues.py b/.github/scripts/close_duplicate_issues.py
deleted file mode 100755
index ec522af4f88..00000000000
--- a/.github/scripts/close_duplicate_issues.py
+++ /dev/null
@@ -1,230 +0,0 @@
-#!/usr/bin/env python3
-"""
-Detect and close duplicate GitHub issues using title similarity.
-
-Modes:
-  --scan           Compare all open issues against each other (batch)
-  --issue-number N Check a single issue against older open issues
-
-Requires the `gh` CLI to be authenticated.
-"""
-
-import argparse
-import difflib
-import json
-import re
-import subprocess
-import sys
-
-
-def normalize_title(title: str) -> str:
-    """Strip common prefixes, lowercase, and collapse whitespace."""
-    title = re.sub(
-        r"^\[?(bug|feature request|enhancement|question|docs)[:\]]?\s*",
-        "",
-        title,
-        flags=re.IGNORECASE,
-    )
-    return " ".join(title.lower().split())
-
-
-def gh(*args: str) -> str:
-    """Run a gh CLI command and return stdout."""
-    result = subprocess.run(
-        ["gh", *args],
-        capture_output=True,
-        text=True,
-        check=True,
-    )
-    return result.stdout
-
-
-def fetch_open_issues(repo: str | None) -> list[dict]:
-    """Fetch all open issues (excluding PRs) via gh api --paginate."""
-    if repo:
-        endpoint = (
-            f"repos/{repo}/issues?state=open&per_page=100&sort=created&direction=asc"
-        )
-    else:
-        endpoint = "repos/{owner}/{repo}/issues?state=open&per_page=100&sort=created&direction=asc"
-    cmd = ["api", "--paginate", endpoint]
-
-    raw = gh(*cmd)
-    # gh --paginate concatenates JSON arrays, so we may get multiple arrays
-    issues = []
-    for line in raw.strip().splitlines():
-        line = line.strip()
-        if not line:
-            continue
-        parsed = json.loads(line)
-        if isinstance(parsed, list):
-            issues.extend(parsed)
-        else:
-            issues.append(parsed)
-
-    # Filter out pull requests (they also appear in the issues endpoint)
-    return [i for i in issues if "pull_request" not in i]
-
-
-def close_as_duplicate(
-    issue_number: int, duplicate_of: int, repo: str | None, dry_run: bool
-) -> None:
-    """Close an issue as duplicate of another, adding a comment and label."""
-    repo_args = ["--repo", repo] if repo else []
-
-    if dry_run:
-        print(
-            f"  [DRY RUN] Would close #{issue_number} as duplicate of #{duplicate_of}"
-        )
-        return
-
-    # Add comment
-    comment_body = (
-        f"Closing as duplicate of #{duplicate_of}.\n\n"
-        "If you believe this is not a duplicate, please reopen and add context "
-        "explaining how this differs."
-    )
-    gh("issue", "comment", str(issue_number), "--body", comment_body, *repo_args)
-
-    # Add label
-    gh("issue", "edit", str(issue_number), "--add-label", "duplicate", *repo_args)
-
-    # Close with not_planned reason
-    gh(
-        "api",
-        f"repos/{repo or '{owner}/{repo}'}/issues/{issue_number}",
-        "-X",
-        "PATCH",
-        "-f",
-        "state=closed",
-        "-f",
-        "state_reason=not_planned",
-    )
-
-    print(f"  Closed #{issue_number} as duplicate of #{duplicate_of}")
-
-
-def find_duplicate(
-    issue: dict, candidates: list[dict], threshold: float
-) -> dict | None:
-    """Return the first candidate whose normalized title is above threshold."""
-    norm = normalize_title(issue["title"])
-    for candidate in candidates:
-        if candidate["number"] == issue["number"]:
-            continue
-        cand_norm = normalize_title(candidate["title"])
-        ratio = difflib.SequenceMatcher(None, norm, cand_norm).ratio()
-        if ratio >= threshold:
-            return candidate
-    return None
-
-
-def scan_all(
-    issues: list[dict], threshold: float, repo: str | None, dry_run: bool
-) -> int:
-    """Compare every issue against all older issues. Returns count of duplicates found."""
-    # Sort oldest first
-    issues.sort(key=lambda i: i["number"])
-    closed_count = 0
-
-    for idx, issue in enumerate(issues):
-        older = issues[:idx]
-        if not older:
-            continue
-        dup = find_duplicate(issue, older, threshold)
-        if dup:
-            ratio = difflib.SequenceMatcher(
-                None,
-                normalize_title(issue["title"]),
-                normalize_title(dup["title"]),
-            ).ratio()
-            print(
-                f"#{issue['number']}: \"{issue['title']}\"\n"
-                f"  -> duplicate of #{dup['number']}: \"{dup['title']}\" "
-                f"({ratio:.0%} similar)"
-            )
-            close_as_duplicate(issue["number"], dup["number"], repo, dry_run)
-            closed_count += 1
-
-    return closed_count
-
-
-def check_single(
-    issue_number: int,
-    issues: list[dict],
-    threshold: float,
-    repo: str | None,
-    dry_run: bool,
-) -> bool:
-    """Check a single issue against all older open issues. Returns True if duplicate found."""
-    target = None
-    for i in issues:
-        if i["number"] == issue_number:
-            target = i
-            break
-
-    if target is None:
-        print(f"Issue #{issue_number} not found among open issues.")
-        return False
-
-    older = [i for i in issues if i["number"] < issue_number]
-    dup = find_duplicate(target, older, threshold)
-    if dup:
-        ratio = difflib.SequenceMatcher(
-            None,
-            normalize_title(target["title"]),
-            normalize_title(dup["title"]),
-        ).ratio()
-        print(
-            f"#{target['number']}: \"{target['title']}\"\n"
-            f"  -> duplicate of #{dup['number']}: \"{dup['title']}\" "
-            f"({ratio:.0%} similar)"
-        )
-        close_as_duplicate(issue_number, dup["number"], repo, dry_run)
-        return True
-
-    print(f"#{issue_number}: no duplicate found above threshold {threshold}")
-    return False
-
-
-def main() -> None:
-    parser = argparse.ArgumentParser(
-        description="Detect and close duplicate GitHub issues"
-    )
-    mode = parser.add_mutually_exclusive_group(required=True)
-    mode.add_argument("--scan", action="store_true", help="Scan all open issues")
-    mode.add_argument("--issue-number", type=int, help="Check a single issue number")
-    parser.add_argument(
-        "--threshold", type=float, default=0.85, help="Similarity threshold (0-1)"
-    )
-    parser.add_argument(
-        "--close",
-        action="store_true",
-        help="Actually close duplicates (default is dry-run)",
-    )
-    parser.add_argument(
-        "--repo", type=str, help="Repository (owner/repo). Auto-detected if omitted."
-    )
-    args = parser.parse_args()
-
-    dry_run = not args.close
-
-    if dry_run:
-        print("=== DRY RUN MODE (pass --close to actually close issues) ===\n")
-
-    print("Fetching open issues...")
-    issues = fetch_open_issues(args.repo)
-    print(f"Found {len(issues)} open issues.\n")
-
-    if args.scan:
-        count = scan_all(issues, args.threshold, args.repo, dry_run)
-        print(f"\nTotal duplicates {'found' if dry_run else 'closed'}: {count}")
-    else:
-        found = check_single(
-            args.issue_number, issues, args.threshold, args.repo, dry_run
-        )
-        sys.exit(0 if found else 0)  # Always exit 0; finding no dup is not an error
-
-
-if __name__ == "__main__":
-    main()
diff --git a/.github/workflows/auto-close-duplicates.yml b/.github/workflows/auto-close-duplicates.yml
new file mode 100644
index 00000000000..d8256917805
--- /dev/null
+++ b/.github/workflows/auto-close-duplicates.yml
@@ -0,0 +1,69 @@
+name: Auto-close duplicate issues
+
+on:
+  schedule:
+    - cron: "0 9 * * *"
+  workflow_dispatch:
+    inputs:
+      dry_run:
+        description: Log which issues would close without closing anything
+        type: boolean
+        default: true
+      grace_period_days:
+        description: Days a duplicate notice must go unanswered before the close
+        type: number
+        default: 3
+  pull_request:
+    paths:
+      - .github/workflows/auto-close-duplicates.yml
+      - scripts/auto-close-duplicates.ts
+      - scripts/auto-close-duplicates.test.ts
+
+permissions: {}
+
+jobs:
+  test:
+    if: github.event_name == 'pull_request'
+    runs-on: ubuntu-latest
+    timeout-minutes: 5
+    permissions:
+      contents: read
+    steps:
+      - name: Checkout repository
+        uses: actions/checkout@08eba0b27e820071cde6df949e0beb9ba4906955 # v4.3.0
+        with:
+          persist-credentials: false
+
+      - name: Setup Bun
+        uses: oven-sh/setup-bun@0c5077e51419868618aeaa5fe8019c62421857d6 # v2.2.0
+        with:
+          bun-version: "1.4.0"
+
+      - name: Test the sweep
+        run: bun test scripts/auto-close-duplicates.test.ts
+
+  sweep:
+    if: github.event_name != 'pull_request' && github.repository == 'BerriAI/litellm'
+    runs-on: ubuntu-latest
+    timeout-minutes: 10
+    permissions:
+      contents: read
+      issues: write
+    steps:
+      - name: Checkout repository
+        uses: actions/checkout@08eba0b27e820071cde6df949e0beb9ba4906955 # v4.3.0
+        with:
+          persist-credentials: false
+
+      - name: Setup Bun
+        uses: oven-sh/setup-bun@0c5077e51419868618aeaa5fe8019c62421857d6 # v2.2.0
+        with:
+          # Exact version, never latest: the next step holds an issues: write token
+          bun-version: "1.4.0"
+
+      - name: Close unanswered duplicates, reopen ones the reporter answered
+        run: bun run scripts/auto-close-duplicates.ts
+        env:
+          GITHUB_TOKEN: ${{ secrets.GITHUB_TOKEN }}
+          DRY_RUN: ${{ inputs.dry_run == true }}
+          GRACE_PERIOD_DAYS: ${{ inputs.grace_period_days }}
diff --git a/.github/workflows/check_duplicate_issues.yml b/.github/workflows/check_duplicate_issues.yml
index 78198b2c7bb..41ec43a1d9b 100644
--- a/.github/workflows/check_duplicate_issues.yml
+++ b/.github/workflows/check_duplicate_issues.yml
@@ -1,12 +1,19 @@
 name: Check Duplicate Issues
 
+# Flagging only. "Auto-close duplicate issues" closes a flagged issue 3 days later,
+# and only when its title is identical to an older open issue and nobody replied.
+# The HTML marker below is the handshake between the two, so keep it in the template.
+
 on:
   issues:
     types: [opened, edited]
 
+permissions: {}
+
 jobs:
   check-duplicate:
     runs-on: ubuntu-latest
+    timeout-minutes: 5
     permissions:
       issues: write
       contents: read
@@ -19,35 +26,12 @@ jobs:
           threshold: 0.6
           reaction: eyes
           comment: |
-            **⚠️ Potential duplicate detected**
+            <!-- litellm:potential-duplicate candidates={{#issues}}{{number}},{{/issues}} -->
+            **Potential duplicate detected**
 
-            This issue appears similar to existing issue(s):
+            This looks similar to:
             {{#issues}}
-            - [#{{number}}]({{html_url}}) - {{title}} ({{accuracy}}% similar)
+            - #{{number}} - {{title}}
             {{/issues}}
 
-            Please review the linked issue(s) to see if they address your concern. If this is not a duplicate, please provide additional context to help us understand the difference.
-
-      - name: Checkout close script
-        if: github.event.action == 'opened'
-        uses: actions/checkout@08eba0b27e820071cde6df949e0beb9ba4906955 # v4.3.0
-        with:
-          sparse-checkout: .github/scripts
-          persist-credentials: false
-
-      - name: Set up Python
-        if: github.event.action == 'opened'
-        uses: actions/setup-python@a26af69be951a213d495a4c3e4e4022e16d87065 # v5.6.0
-        with:
-          python-version: "3.12"
-
-      - name: Auto-close if high-confidence duplicate
-        if: github.event.action == 'opened'
-        env:
-          GH_TOKEN: ${{ secrets.GITHUB_TOKEN }}
-        run: |
-          python3 .github/scripts/close_duplicate_issues.py \
-            --issue-number ${{ github.event.issue.number }} \
-            --repo ${{ github.repository }} \
-            --threshold 0.85 \
-            --close
+            If this is a duplicate, add a thumbs-up reaction to the existing issue and follow along there. When the title is identical to an older open issue, this issue closes automatically in 3 days unless someone responds. If it is not a duplicate, comment here or add a thumbs-down reaction to this comment and it stays open.
diff --git a/litellm/model_prices_and_context_window_backup.json b/litellm/model_prices_and_context_window_backup.json
index 05c1cfd3179..7071eaa0807 100644
--- a/litellm/model_prices_and_context_window_backup.json
+++ b/litellm/model_prices_and_context_window_backup.json
@@ -3726,7 +3726,7 @@
         "output_cost_per_token": 0,
         "litellm_provider": "azure_ai",
         "mode": "chat",
-        "source": "https://azure.microsoft.com/en-us/pricing/details/ai-services/",
+        "source": "https://azure.microsoft.com/en-us/pricing/details/ai-foundry-models/aoai/",
         "comment": "Flat cost of $0.14 per M input tokens for Azure AI Foundry Model Router infrastructure. Use pattern: azure_ai/model_router/<deployment-name> where deployment-name is your Azure deployment (e.g., azure-model-router)"
     },
     "azure/eu/gpt-4o-2024-08-06": {
@@ -5328,7 +5328,7 @@
         "input_cost_per_second": 0.0002833333333333333,
         "litellm_provider": "azure",
         "mode": "audio_transcription",
-        "source": "https://learn.microsoft.com/en-us/azure/ai-foundry/openai/concepts/gpt-realtime-whisper",
+        "source": "https://learn.microsoft.com/en-us/azure/foundry/openai/concepts/gpt-realtime-whisper",
         "supported_endpoints": [
             "/v1/realtime",
             "/v1/realtime/transcription_sessions"
@@ -9107,7 +9107,7 @@
         "mode": "embedding",
         "output_cost_per_token": 0.0,
         "output_vector_size": 1024,
-        "source": "https://azuremarketplace.microsoft.com/en-us/marketplace/apps/cohere.cohere-embed-v3-english-offer?tab=PlansAndPrice",
+        "source": "https://marketplace.microsoft.com/en-us/marketplace/apps/cohere.cohere-embed-v3-english-offer?tab=PlansAndPrice",
         "supports_embedding_image_input": true
     },
     "azure_ai/Cohere-embed-v3-multilingual": {
@@ -9118,7 +9118,7 @@
         "mode": "embedding",
         "output_cost_per_token": 0.0,
         "output_vector_size": 1024,
-        "source": "https://azuremarketplace.microsoft.com/en-us/marketplace/apps/cohere.cohere-embed-v3-english-offer?tab=PlansAndPrice",
+        "source": "https://marketplace.microsoft.com/en-us/marketplace/apps/cohere.cohere-embed-v3-english-offer?tab=PlansAndPrice",
         "supports_embedding_image_input": true
     },
     "azure_ai/FLUX-1.1-pro": {
@@ -9134,7 +9134,7 @@
         "litellm_provider": "azure_ai",
         "mode": "image_generation",
         "output_cost_per_image": 0.04,
-        "source": "https://azuremarketplace.microsoft.com/pt-br/marketplace/apps/cohere.cohere-embed-4-offer?tab=PlansAndPrice",
+        "source": "https://marketplace.microsoft.com/pt-br/marketplace/apps/cohere.cohere-embed-4-offer?tab=PlansAndPrice",
         "supported_endpoints": [
             "/v1/images/generations"
         ]
@@ -9467,7 +9467,7 @@
         "max_tokens": 2048,
         "mode": "chat",
         "output_cost_per_token": 3.7e-07,
-        "source": "https://azuremarketplace.microsoft.com/en/marketplace/apps/metagenai.meta-llama-3-2-11b-vision-instruct-offer?tab=Overview",
+        "source": "https://marketplace.microsoft.com/en/marketplace/apps/metagenai.meta-llama-3-2-11b-vision-instruct-offer?tab=Overview",
         "supports_function_calling": true,
         "supports_tool_choice": true,
         "supports_vision": true
@@ -9481,7 +9481,7 @@
         "max_tokens": 2048,
         "mode": "chat",
         "output_cost_per_token": 2.04e-06,
-        "source": "https://azuremarketplace.microsoft.com/en/marketplace/apps/metagenai.meta-llama-3-2-90b-vision-instruct-offer?tab=Overview",
+        "source": "https://marketplace.microsoft.com/en/marketplace/apps/metagenai.meta-llama-3-2-90b-vision-instruct-offer?tab=Overview",
         "supports_function_calling": true,
         "supports_tool_choice": true,
         "supports_vision": true
@@ -9494,7 +9494,7 @@
         "max_tokens": 2048,
         "mode": "chat",
         "output_cost_per_token": 7.1e-07,
-        "source": "https://azuremarketplace.microsoft.com/en/marketplace/apps/metagenai.llama-3-3-70b-instruct-offer?tab=Overview",
+        "source": "https://marketplace.microsoft.com/en/marketplace/apps/metagenai.llama-3-3-70b-instruct-offer?tab=Overview",
         "supports_function_calling": true,
         "supports_tool_choice": true
     },
@@ -9543,7 +9543,7 @@
         "max_tokens": 2048,
         "mode": "chat",
         "output_cost_per_token": 1.6e-05,
-        "source": "https://azuremarketplace.microsoft.com/en-us/marketplace/apps/metagenai.meta-llama-3-1-405b-instruct-offer?tab=PlansAndPrice",
+        "source": "https://marketplace.microsoft.com/en-us/marketplace/apps/metagenai.meta-llama-3-1-405b-instruct-offer?tab=PlansAndPrice",
         "supports_tool_choice": true
     },
     "azure_ai/Meta-Llama-3.1-70B-Instruct": {
@@ -9554,7 +9554,7 @@
         "max_tokens": 2048,
         "mode": "chat",
         "output_cost_per_token": 3.54e-06,
-        "source": "https://azuremarketplace.microsoft.com/en-us/marketplace/apps/metagenai.meta-llama-3-1-70b-instruct-offer?tab=PlansAndPrice",
+        "source": "https://marketplace.microsoft.com/en-us/marketplace/apps/metagenai.meta-llama-3-1-70b-instruct-offer?tab=PlansAndPrice",
         "supports_tool_choice": true
     },
     "azure_ai/Meta-Llama-3.1-8B-Instruct": {
@@ -9566,7 +9566,7 @@
         "max_tokens": 2048,
         "mode": "chat",
         "output_cost_per_token": 6.1e-07,
-        "source": "https://azuremarketplace.microsoft.com/en-us/marketplace/apps/metagenai.meta-llama-3-1-8b-instruct-offer?tab=PlansAndPrice",
+        "source": "https://marketplace.microsoft.com/en-us/marketplace/apps/metagenai.meta-llama-3-1-8b-instruct-offer?tab=PlansAndPrice",
         "supports_tool_choice": true
     },
     "azure_ai/Phi-3-medium-128k-instruct": {
@@ -9756,7 +9756,7 @@
         "supported_endpoints": [
             "/v1/ocr"
         ],
-        "source": "https://azure.microsoft.com/en-us/pricing/details/ai-foundry-models/"
+        "source": "https://ai.azure.com/catalog/models/mistral-document-ai-2512"
     },
     "azure_ai/doc-intelligence/prebuilt-read": {
         "litellm_provider": "azure_ai",
@@ -9967,7 +9967,7 @@
         "mode": "embedding",
         "output_cost_per_token": 0.0,
         "output_vector_size": 3072,
-        "source": "https://azuremarketplace.microsoft.com/pt-br/marketplace/apps/cohere.cohere-embed-4-offer?tab=PlansAndPrice",
+        "source": "https://marketplace.microsoft.com/pt-br/marketplace/apps/cohere.cohere-embed-4-offer?tab=PlansAndPrice",
         "supported_endpoints": [
             "/v1/embeddings"
         ],
@@ -10151,7 +10151,7 @@
         "max_tokens": 8192,
         "mode": "chat",
         "output_cost_per_token": 0.00971,
-        "source": "https://azure.microsoft.com/en-us/products/ai-services/ai-foundry/models/jais-30b-chat"
+        "source": "https://ai.azure.com/catalog/models/jais-30b-chat"
     },
     "azure_ai/jamba-instruct": {
         "input_cost_per_token": 5e-07,
@@ -10208,7 +10208,7 @@
         "max_tokens": 4096,
         "mode": "chat",
         "output_cost_per_token": 4e-08,
-        "source": "https://azuremarketplace.microsoft.com/en/marketplace/apps/000-000.ministral-3b-2410-offer?tab=Overview",
+        "source": "https://marketplace.microsoft.com/en/marketplace/apps/000-000.ministral-3b-2410-offer?tab=Overview",
         "supports_function_calling": true,
         "supports_tool_choice": true
     },
@@ -10231,7 +10231,7 @@
         "max_tokens": 4096,
         "mode": "chat",
         "output_cost_per_token": 6e-06,
-        "source": "https://azuremarketplace.microsoft.com/en/marketplace/apps/000-000.mistral-ai-large-2407-offer?tab=Overview",
+        "source": "https://marketplace.microsoft.com/en/marketplace/apps/000-000.mistral-ai-large-2407-offer?tab=Overview",
         "supports_function_calling": true,
         "supports_tool_choice": true
     },
@@ -10243,7 +10243,7 @@
         "max_tokens": 4096,
         "mode": "chat",
         "output_cost_per_token": 6e-06,
-        "source": "https://azuremarketplace.microsoft.com/en/marketplace/apps/000-000.mistral-ai-large-2407-offer?tab=Overview",
+        "source": "https://marketplace.microsoft.com/en/marketplace/apps/000-000.mistral-ai-large-2407-offer?tab=Overview",
         "supports_function_calling": true,
         "supports_tool_choice": true
     },
@@ -10280,7 +10280,7 @@
         "max_tokens": 4096,
         "mode": "chat",
         "output_cost_per_token": 1.5e-07,
-        "source": "https://azuremarketplace.microsoft.com/en/marketplace/apps/000-000.mistral-nemo-12b-2407?tab=PlansAndPrice",
+        "source": "https://marketplace.microsoft.com/en/marketplace/apps/000-000.mistral-nemo-12b-2407?tab=PlansAndPrice",
         "supports_function_calling": true
     },
     "azure_ai/mistral-small": {
@@ -12315,7 +12315,8 @@
         "output_cost_per_token": 1e-05,
         "supported_endpoints": [
             "/v1/audio/transcriptions"
-        ]
+        ],
+        "deprecation_date": "2027-02-26"
     },
     "claude-haiku-4-5-20251001": {
         "deprecation_date": "2026-10-15",
@@ -17767,7 +17768,8 @@
         "supports_response_schema": true,
         "supports_tool_choice": true,
         "supports_vision": true,
-        "prompt_cache_min_tokens": 1024
+        "prompt_cache_min_tokens": 1024,
+        "deprecation_date": "2027-01-08"
     },
     "eu.anthropic.claude-opus-4-20250514-v1:0": {
         "cache_creation_input_token_cost": 1.875e-05,
@@ -22843,7 +22845,7 @@
         "output_cost_per_reasoning_token": 3e-06,
         "output_cost_per_token": 3e-06,
         "rpm": 2000,
-        "source": "https://ai.google.dev/pricing/gemini-3",
+        "source": "https://ai.google.dev/gemini-api/docs/pricing",
         "supported_endpoints": [
             "/v1/chat/completions",
             "/v1/completions",
@@ -22897,7 +22899,7 @@
         "output_cost_per_reasoning_token": 9e-06,
         "output_cost_per_token": 9e-06,
         "rpm": 2000,
-        "source": "https://ai.google.dev/pricing/gemini-3",
+        "source": "https://ai.google.dev/gemini-api/docs/pricing",
         "supported_endpoints": [
             "/v1/chat/completions",
             "/v1/completions",
@@ -22962,7 +22964,7 @@
         "output_cost_per_token_batches": 1.875e-06,
         "output_cost_per_token_flex": 1.875e-06,
         "rpm": 2000,
-        "source": "https://ai.google.dev/pricing/gemini-3",
+        "source": "https://ai.google.dev/gemini-api/docs/pricing",
         "supported_endpoints": [
             "/v1/chat/completions",
             "/v1/completions",
@@ -23021,7 +23023,7 @@
         "output_cost_per_token_batches": 1.875e-06,
         "output_cost_per_token_flex": 1.875e-06,
         "rpm": 2000,
-        "source": "https://ai.google.dev/pricing/gemini-3",
+        "source": "https://ai.google.dev/gemini-api/docs/pricing",
         "supported_endpoints": [
             "/v1/chat/completions",
             "/v1/completions",
@@ -23226,7 +23228,7 @@
         "mode": "chat",
         "output_cost_per_reasoning_token": 3e-06,
         "output_cost_per_token": 3e-06,
-        "source": "https://ai.google.dev/pricing/gemini-3",
+        "source": "https://ai.google.dev/gemini-api/docs/pricing",
         "supported_endpoints": [
             "/v1/chat/completions",
             "/v1/completions",
@@ -23310,7 +23312,7 @@
         "mode": "chat",
         "output_cost_per_reasoning_token": 9e-06,
         "output_cost_per_token": 9e-06,
-        "source": "https://ai.google.dev/pricing/gemini-3",
+        "source": "https://ai.google.dev/gemini-api/docs/pricing",
         "supported_endpoints": [
             "/v1/chat/completions",
             "/v1/completions",
@@ -23373,7 +23375,7 @@
         "output_cost_per_token": 3.75e-06,
         "output_cost_per_token_batches": 1.875e-06,
         "output_cost_per_token_flex": 1.875e-06,
-        "source": "https://ai.google.dev/pricing/gemini-3",
+        "source": "https://ai.google.dev/gemini-api/docs/pricing",
         "supported_endpoints": [
             "/v1/chat/completions",
             "/v1/completions",
@@ -23430,7 +23432,7 @@
         "output_cost_per_token": 3.75e-06,
         "output_cost_per_token_batches": 1.875e-06,
         "output_cost_per_token_flex": 1.875e-06,
-        "source": "https://ai.google.dev/pricing/gemini-3",
+        "source": "https://ai.google.dev/gemini-api/docs/pricing",
         "supported_endpoints": [
             "/v1/chat/completions",
             "/v1/completions",
@@ -25327,7 +25329,7 @@
         "supports_vision": true
     },
     "gpt-4o-audio-preview": {
-        "deprecation_date": "2027-01-20",
+        "deprecation_date": "2026-05-07",
         "input_cost_per_audio_token": 4e-05,
         "input_cost_per_token": 2.5e-06,
         "litellm_provider": "openai",
@@ -25650,7 +25652,7 @@
         "supports_vision": true
     },
     "gpt-4o-mini-audio-preview": {
-        "deprecation_date": "2027-01-20",
+        "deprecation_date": "2026-05-07",
         "input_cost_per_audio_token": 1e-05,
         "input_cost_per_token": 1.5e-07,
         "litellm_provider": "openai",
@@ -25688,7 +25690,7 @@
     "gpt-4o-mini-realtime-preview": {
         "cache_creation_input_audio_token_cost": 3e-07,
         "cache_read_input_token_cost": 3e-07,
-        "deprecation_date": "2027-01-20",
+        "deprecation_date": "2026-05-07",
         "input_cost_per_audio_token": 1e-05,
         "input_cost_per_token": 6e-07,
         "litellm_provider": "openai",
@@ -25787,7 +25789,8 @@
         "output_cost_per_token": 5e-06,
         "supported_endpoints": [
             "/v1/audio/transcriptions"
-        ]
+        ],
+        "deprecation_date": "2027-02-26"
     },
     "gpt-4o-mini-tts": {
         "input_cost_per_token": 2.5e-06,
@@ -25809,7 +25812,7 @@
     },
     "gpt-4o-realtime-preview": {
         "cache_read_input_token_cost": 2.5e-06,
-        "deprecation_date": "2027-01-20",
+        "deprecation_date": "2026-05-07",
         "input_cost_per_audio_token": 4e-05,
         "input_cost_per_token": 5e-06,
         "litellm_provider": "openai",
@@ -25828,7 +25831,7 @@
     },
     "gpt-4o-realtime-preview-2024-12-17": {
         "cache_read_input_token_cost": 2.5e-06,
-        "deprecation_date": "2027-01-20",
+        "deprecation_date": "2026-05-07",
         "input_cost_per_audio_token": 4e-05,
         "input_cost_per_token": 5e-06,
         "litellm_provider": "openai",
@@ -25847,7 +25850,7 @@
     },
     "gpt-4o-realtime-preview-2025-06-03": {
         "cache_read_input_token_cost": 2.5e-06,
-        "deprecation_date": "2027-01-20",
+        "deprecation_date": "2026-05-07",
         "input_cost_per_audio_token": 4e-05,
         "input_cost_per_token": 5e-06,
         "litellm_provider": "openai",
@@ -25926,7 +25929,8 @@
         "output_cost_per_token": 1e-05,
         "supported_endpoints": [
             "/v1/audio/transcriptions"
-        ]
+        ],
+        "deprecation_date": "2027-02-26"
     },
     "gpt-image-1.5": {
         "cache_read_input_token_cost": 1.25e-06,
@@ -27077,7 +27081,7 @@
         "supports_tool_choice": true,
         "supports_vision": true,
         "supports_web_search": true,
-        "source": "https://platform.openai.com/docs/models/gpt-5.6-cyber",
+        "source": "https://developers.openai.com/api/docs/models/gpt-5.6-cyber",
         "supports_computer_use": true,
         "supports_parallel_function_calling": true
     },
@@ -27116,7 +27120,7 @@
         "supports_tool_choice": true,
         "supports_vision": true,
         "supports_web_search": true,
-        "source": "https://platform.openai.com/docs/models/daybreak-red-latest",
+        "source": "https://developers.openai.com/api/docs/models/daybreak-red-latest",
         "supports_computer_use": true,
         "supports_parallel_function_calling": true
     },
@@ -27156,7 +27160,7 @@
         "supports_tool_choice": true,
         "supports_vision": true,
         "supports_web_search": true,
-        "source": "https://platform.openai.com/docs/models/daybreak-blue-latest",
+        "source": "https://developers.openai.com/api/docs/models/daybreak-blue-latest",
         "supports_parallel_function_calling": true
     },
     "chat-latest": {
@@ -27168,7 +27172,7 @@
         "max_tokens": 128000,
         "mode": "chat",
         "output_cost_per_token": 3e-05,
-        "source": "https://platform.openai.com/docs/models/chat-latest",
+        "source": "https://developers.openai.com/api/docs/models/chat-latest",
         "supported_endpoints": [
             "/v1/chat/completions",
             "/v1/responses"
@@ -30463,7 +30467,7 @@
             "search_context_size_low": 0.0025,
             "search_context_size_medium": 0.0025
         },
-        "source": "https://dev.meta.ai/docs/getting-started/pricing-rate-limits",
+        "source": "https://ai.developer.meta.com/docs/pricing-rate-limits",
         "supported_endpoints": [
             "/v1/chat/completions",
             "/v1/responses",
@@ -30504,7 +30508,7 @@
             "search_context_size_low": 0.0025,
             "search_context_size_medium": 0.0025
         },
-        "source": "https://dev.meta.ai/docs/getting-started/pricing-rate-limits",
+        "source": "https://ai.developer.meta.com/docs/pricing-rate-limits",
         "supported_endpoints": [
             "/v1/chat/completions",
             "/v1/responses",
@@ -30545,7 +30549,7 @@
             "search_context_size_low": 0.0025,
             "search_context_size_medium": 0.0025
         },
-        "source": "https://dev.meta.ai/docs/getting-started/pricing-rate-limits",
+        "source": "https://ai.developer.meta.com/docs/pricing-rate-limits",
         "supported_endpoints": [
             "/v1/chat/completions",
             "/v1/responses",
@@ -30578,7 +30582,7 @@
         "max_output_tokens": 4028,
         "max_tokens": 4028,
         "mode": "chat",
-        "source": "https://llama.developer.meta.com/docs/models",
+        "source": "https://ai.developer.meta.com/docs/models",
         "supported_modalities": [
             "text"
         ],
@@ -30594,7 +30598,7 @@
         "max_output_tokens": 4028,
         "max_tokens": 4028,
         "mode": "chat",
-        "source": "https://llama.developer.meta.com/docs/models",
+        "source": "https://ai.developer.meta.com/docs/models",
         "supported_modalities": [
             "text"
         ],
@@ -30610,7 +30614,7 @@
         "max_output_tokens": 4028,
         "max_tokens": 4028,
         "mode": "chat",
-        "source": "https://llama.developer.meta.com/docs/models",
+        "source": "https://ai.developer.meta.com/docs/models",
         "supported_modalities": [
             "text",
             "image"
@@ -30627,7 +30631,7 @@
         "max_output_tokens": 4028,
         "max_tokens": 4028,
         "mode": "chat",
-        "source": "https://llama.developer.meta.com/docs/models",
+        "source": "https://ai.developer.meta.com/docs/models",
         "supported_modalities": [
             "text",
             "image"
@@ -32572,7 +32576,7 @@
         "mode": "chat",
         "supports_function_calling": true,
         "supports_reasoning": true,
-        "source": "https://nebius.com/prices-ai-studio"
+        "source": "https://nebius.com/prices"
     },
     "nebius/deepseek-ai/DeepSeek-R1-0528": {
         "max_tokens": 164000,
@@ -32584,7 +32588,7 @@
         "mode": "chat",
         "supports_function_calling": true,
         "supports_reasoning": true,
-        "source": "https://nebius.com/prices-ai-studio"
+        "source": "https://nebius.com/prices"
     },
     "nebius/deepseek-ai/DeepSeek-R1-Distill-Llama-70B": {
         "max_tokens": 128000,
@@ -32595,7 +32599,7 @@
         "litellm_provider": "nebius",
         "mode": "chat",
         "supports_function_calling": true,
-        "source": "https://nebius.com/prices-ai-studio"
+        "source": "https://nebius.com/prices"
     },
     "nebius/deepseek-ai/DeepSeek-V3": {
         "max_tokens": 128000,
@@ -32606,7 +32610,7 @@
         "litellm_provider": "nebius",
         "mode": "chat",
         "supports_function_calling": true,
-        "source": "https://nebius.com/prices-ai-studio"
+        "source": "https://nebius.com/prices"
     },
     "nebius/deepseek-ai/DeepSeek-V3-0324": {
         "max_tokens": 128000,
@@ -32617,7 +32621,7 @@
         "litellm_provider": "nebius",
         "mode": "chat",
         "supports_function_calling": true,
-        "source": "https://nebius.com/prices-ai-studio"
+        "source": "https://nebius.com/prices"
     },
     "nebius/google/gemma-3-27b-it": {
         "max_tokens": 128000,
@@ -32629,7 +32633,7 @@
         "mode": "chat",
         "supports_function_calling": true,
         "supports_vision": true,
-        "source": "https://nebius.com/prices-ai-studio"
+        "source": "https://nebius.com/prices"
     },
     "nebius/meta-llama/Llama-3.3-70B-Instruct": {
         "max_tokens": 128000,
@@ -32640,7 +32644,7 @@
         "litellm_provider": "nebius",
         "mode": "chat",
         "supports_function_calling": true,
-        "source": "https://nebius.com/prices-ai-studio"
+        "source": "https://nebius.com/prices"
     },
     "nebius/meta-llama/Llama-Guard-3-8B": {
         "max_tokens": 128000,
@@ -32650,7 +32654,7 @@
         "output_cost_per_token": 6e-08,
         "litellm_provider": "nebius",
         "mode": "chat",
-        "source": "https://nebius.com/prices-ai-studio"
+        "source": "https://nebius.com/prices"
     },
     "nebius/meta-llama/Meta-Llama-3.1-8B-Instruct": {
         "max_tokens": 128000,
@@ -32661,7 +32665,7 @@
         "litellm_provider": "nebius",
         "mode": "chat",
         "supports_function_calling": true,
-        "source": "https://nebius.com/prices-ai-studio"
+        "source": "https://nebius.com/prices"
     },
     "nebius/meta-llama/Meta-Llama-3.1-70B-Instruct": {
         "max_tokens": 128000,
@@ -32672,7 +32676,7 @@
         "litellm_provider": "nebius",
         "mode": "chat",
         "supports_function_calling": true,
-        "source": "https://nebius.com/prices-ai-studio"
+        "source": "https://nebius.com/prices"
     },
     "nebius/meta-llama/Meta-Llama-3.1-405B-Instruct": {
         "max_tokens": 128000,
@@ -32683,7 +32687,7 @@
         "litellm_provider": "nebius",
         "mode": "chat",
         "supports_function_calling": true,
-        "source": "https://nebius.com/prices-ai-studio"
+        "source": "https://nebius.com/prices"
     },
     "nebius/mistralai/Mistral-Nemo-Instruct-2407": {
         "max_tokens": 128000,
@@ -32694,7 +32698,7 @@
         "litellm_provider": "nebius",
         "mode": "chat",
         "supports_function_calling": true,
-        "source": "https://nebius.com/prices-ai-studio"
+        "source": "https://nebius.com/prices"
     },
     "nebius/NousResearch/Hermes-3-Llama-3.1-405B": {
         "max_tokens": 128000,
@@ -32705,7 +32709,7 @@
         "litellm_provider": "nebius",
         "mode": "chat",
         "supports_function_calling": true,
-        "source": "https://nebius.com/prices-ai-studio"
+        "source": "https://nebius.com/prices"
     },
     "nebius/nvidia/Llama-3.1-Nemotron-Ultra-253B-v1": {
         "max_tokens": 128000,
@@ -32716,7 +32720,7 @@
         "litellm_provider": "nebius",
         "mode": "chat",
         "supports_function_calling": true,
-        "source": "https://nebius.com/prices-ai-studio"
+        "source": "https://nebius.com/prices"
     },
     "nebius/nvidia/Llama-3.3-Nemotron-Super-49B-v1": {
         "max_tokens": 131072,
@@ -32727,7 +32731,7 @@
         "litellm_provider": "nebius",
         "mode": "chat",
         "supports_function_calling": true,
-        "source": "https://nebius.com/prices-ai-studio"
+        "source": "https://nebius.com/prices"
     },
     "nebius/Qwen/Qwen3-235B-A22B": {
         "max_tokens": 262144,
@@ -32738,7 +32742,7 @@
         "litellm_provider": "nebius",
         "mode": "chat",
         "supports_function_calling": true,
-        "source": "https://nebius.com/prices-ai-studio"
+        "source": "https://nebius.com/prices"
     },
     "nebius/Qwen/Qwen3-32B": {
         "max_tokens": 32768,
@@ -32749,7 +32753,7 @@
         "litellm_provider": "nebius",
         "mode": "chat",
         "supports_function_calling": true,
-        "source": "https://nebius.com/prices-ai-studio"
+        "source": "https://nebius.com/prices"
     },
     "nebius/Qwen/Qwen3-30B-A3B": {
         "max_tokens": 32768,
@@ -32760,7 +32764,7 @@
         "litellm_provider": "nebius",
         "mode": "chat",
         "supports_function_calling": true,
-        "source": "https://nebius.com/prices-ai-studio"
+        "source": "https://nebius.com/prices"
     },
     "nebius/Qwen/Qwen3-14B": {
         "max_tokens": 32768,
@@ -32771,7 +32775,7 @@
         "litellm_provider": "nebius",
         "mode": "chat",
         "supports_function_calling": true,
-        "source": "https://nebius.com/prices-ai-studio"
+        "source": "https://nebius.com/prices"
     },
     "nebius/Qwen/Qwen3-4B": {
         "max_tokens": 32768,
@@ -32782,7 +32786,7 @@
         "litellm_provider": "nebius",
         "mode": "chat",
         "supports_function_calling": true,
-        "source": "https://nebius.com/prices-ai-studio"
+        "source": "https://nebius.com/prices"
     },
     "nebius/Qwen/QwQ-32B": {
         "max_tokens": 32768,
@@ -32794,7 +32798,7 @@
         "mode": "chat",
         "supports_function_calling": true,
         "supports_reasoning": true,
-        "source": "https://nebius.com/prices-ai-studio"
+        "source": "https://nebius.com/prices"
     },
     "nebius/Qwen/Qwen2.5-72B-Instruct": {
         "max_tokens": 128000,
@@ -32805,7 +32809,7 @@
         "litellm_provider": "nebius",
         "mode": "chat",
         "supports_function_calling": true,
-        "source": "https://nebius.com/prices-ai-studio"
+        "source": "https://nebius.com/prices"
     },
     "nebius/Qwen/Qwen2.5-32B-Instruct": {
         "max_tokens": 128000,
@@ -32816,7 +32820,7 @@
         "litellm_provider": "nebius",
         "mode": "chat",
         "supports_function_calling": true,
-        "source": "https://nebius.com/prices-ai-studio"
+        "source": "https://nebius.com/prices"
     },
     "nebius/Qwen/Qwen2.5-Coder-7B": {
         "max_tokens": 32768,
@@ -32827,7 +32831,7 @@
         "litellm_provider": "nebius",
         "mode": "chat",
         "supports_function_calling": true,
-        "source": "https://nebius.com/prices-ai-studio"
+        "source": "https://nebius.com/prices"
     },
     "nebius/Qwen/Qwen2.5-VL-72B-Instruct": {
         "max_tokens": 131072,
@@ -32839,7 +32843,7 @@
         "mode": "chat",
         "supports_function_calling": true,
         "supports_vision": true,
-        "source": "https://nebius.com/prices-ai-studio"
+        "source": "https://nebius.com/prices"
     },
     "nebius/Qwen/Qwen2-VL-72B-Instruct": {
         "max_tokens": 131072,
@@ -32851,7 +32855,7 @@
         "mode": "chat",
         "supports_function_calling": true,
         "supports_vision": true,
-        "source": "https://nebius.com/prices-ai-studio"
+        "source": "https://nebius.com/prices"
     },
     "nebius/Qwen/Qwen2-VL-7B-Instruct": {
         "max_tokens": 131072,
@@ -32862,7 +32866,7 @@
         "litellm_provider": "nebius",
         "mode": "chat",
         "supports_vision": true,
-        "source": "https://nebius.com/prices-ai-studio"
+        "source": "https://nebius.com/prices"
     },
     "nebius/BAAI/bge-en-icl": {
         "max_tokens": 32768,
@@ -32871,7 +32875,7 @@
         "output_cost_per_token": 0.0,
         "litellm_provider": "nebius",
         "mode": "embedding",
-        "source": "https://nebius.com/prices-ai-studio"
+        "source": "https://nebius.com/prices"
     },
     "nebius/BAAI/bge-multilingual-gemma2": {
         "max_tokens": 8192,
@@ -32880,7 +32884,7 @@
         "output_cost_per_token": 0.0,
         "litellm_provider": "nebius",
         "mode": "embedding",
-        "source": "https://nebius.com/prices-ai-studio"
+        "source": "https://nebius.com/prices"
     },
     "nebius/intfloat/e5-mistral-7b-instruct": {
         "max_tokens": 32768,
@@ -32889,7 +32893,7 @@
         "output_cost_per_token": 0.0,
         "litellm_provider": "nebius",
         "mode": "embedding",
-        "source": "https://nebius.com/prices-ai-studio"
+        "source": "https://nebius.com/prices"
     },
     "nvidia.nemotron-nano-12b-v2": {
         "input_cost_per_token": 2e-07,
@@ -33630,7 +33634,7 @@
         "max_tokens": 4000,
         "mode": "chat",
         "output_cost_per_token": 1.56e-06,
-        "source": "https://www.oracle.com/cloud/ai/generative-ai/pricing/",
+        "source": "https://www.oracle.com/artificial-intelligence/enterprise-ai/cost-estimator/",
         "supports_function_calling": true,
         "supports_response_schema": false,
         "supports_native_streaming": true
@@ -33643,7 +33647,7 @@
         "max_tokens": 4000,
         "mode": "chat",
         "output_cost_per_token": 1.56e-06,
-        "source": "https://www.oracle.com/cloud/ai/generative-ai/pricing/",
+        "source": "https://www.oracle.com/artificial-intelligence/enterprise-ai/cost-estimator/",
         "supports_function_calling": true,
         "supports_response_schema": false,
         "supports_native_streaming": true
@@ -33656,7 +33660,7 @@
         "max_tokens": 4000,
         "mode": "chat",
         "output_cost_per_token": 1.56e-06,
-        "source": "https://www.oracle.com/cloud/ai/generative-ai/pricing/",
+        "source": "https://www.oracle.com/artificial-intelligence/enterprise-ai/cost-estimator/",
         "supports_function_calling": true,
         "supports_response_schema": false,
         "supports_native_streaming": true
@@ -33713,7 +33717,7 @@
         "max_tokens": 8192,
         "mode": "chat",
         "output_cost_per_token": 1.56e-06,
-        "source": "https://www.oracle.com/cloud/ai/generative-ai/pricing/",
+        "source": "https://www.oracle.com/artificial-intelligence/enterprise-ai/cost-estimator/",
         "supports_function_calling": true,
         "supports_response_schema": false,
         "supports_native_streaming": true,
@@ -33727,7 +33731,7 @@
         "max_tokens": 8192,
         "mode": "chat",
         "output_cost_per_token": 1.56e-06,
-        "source": "https://www.oracle.com/cloud/ai/generative-ai/pricing/",
+        "source": "https://www.oracle.com/artificial-intelligence/enterprise-ai/cost-estimator/",
         "supports_function_calling": false,
         "supports_response_schema": false,
         "supports_native_streaming": true
@@ -33738,7 +33742,7 @@
         "max_input_tokens": 512,
         "mode": "embedding",
         "output_vector_size": 1024,
-        "source": "https://www.oracle.com/cloud/ai/generative-ai/pricing/",
+        "source": "https://www.oracle.com/artificial-intelligence/enterprise-ai/cost-estimator/",
         "supports_vision": true
     },
     "oci/cohere.command-a-reasoning-08-2025": {
@@ -34635,7 +34639,7 @@
         "max_tokens": 2048,
         "mode": "chat",
         "output_cost_per_token": 2e-07,
-        "source": "https://openrouter.ai/api/v1/models/bytedance/ui-tars-1.5-7b",
+        "source": "https://openrouter.ai/bytedance/ui-tars-1.5-7b",
         "supports_tool_choice": true
     },
     "openrouter/deepseek/deepseek-chat": {
@@ -34870,7 +34874,7 @@
         "output_cost_per_reasoning_token": 3e-06,
         "output_cost_per_token": 3e-06,
         "rpm": 2000,
-        "source": "https://ai.google.dev/pricing/gemini-3",
+        "source": "https://ai.google.dev/gemini-api/docs/pricing",
         "supported_endpoints": [
             "/v1/chat/completions",
             "/v1/completions",
@@ -34911,7 +34915,7 @@
         "output_cost_per_reasoning_token": 1.5e-06,
         "output_cost_per_token": 1.5e-06,
         "rpm": 2000,
-        "source": "https://ai.google.dev/pricing/gemini-3",
+        "source": "https://ai.google.dev/gemini-api/docs/pricing",
         "supported_endpoints": [
             "/v1/chat/completions",
             "/v1/completions",
@@ -35996,7 +36000,7 @@
         "max_tokens": 131000,
         "mode": "chat",
         "output_cost_per_token": 6.7e-07,
-        "source": "https://endpoints.ai.cloud.ovh.net/models/deepseek-r1-distill-llama-70b",
+        "source": "https://www.ovhcloud.com/en/public-cloud/ai-endpoints/catalog/",
         "supports_function_calling": true,
         "supports_reasoning": true,
         "supports_response_schema": true,
@@ -36010,7 +36014,7 @@
         "max_tokens": 131000,
         "mode": "chat",
         "output_cost_per_token": 1e-07,
-        "source": "https://endpoints.ai.cloud.ovh.net/models/llama-3-1-8b-instruct",
+        "source": "https://www.ovhcloud.com/en/public-cloud/ai-endpoints/catalog/",
         "supports_function_calling": true,
         "supports_response_schema": true,
         "supports_tool_choice": true
@@ -36023,7 +36027,7 @@
         "max_tokens": 131000,
         "mode": "chat",
         "output_cost_per_token": 6.7e-07,
-        "source": "https://endpoints.ai.cloud.ovh.net/models/meta-llama-3-1-70b-instruct",
+        "source": "https://www.ovhcloud.com/en/public-cloud/ai-endpoints/catalog/",
         "supports_function_calling": false,
         "supports_response_schema": false,
         "supports_tool_choice": false
@@ -36036,7 +36040,7 @@
         "max_tokens": 131000,
         "mode": "chat",
         "output_cost_per_token": 6.7e-07,
-        "source": "https://endpoints.ai.cloud.ovh.net/models/meta-llama-3-3-70b-instruct",
+        "source": "https://www.ovhcloud.com/en/public-cloud/ai-endpoints/catalog/",
         "supports_function_calling": true,
         "supports_response_schema": true,
         "supports_tool_choice": true
@@ -36049,7 +36053,7 @@
         "max_tokens": 127000,
         "mode": "chat",
         "output_cost_per_token": 1e-07,
-        "source": "https://endpoints.ai.cloud.ovh.net/models/mistral-7b-instruct-v0-3",
+        "source": "https://www.ovhcloud.com/en/public-cloud/ai-endpoints/catalog/",
         "supports_function_calling": true,
         "supports_response_schema": true,
         "supports_tool_choice": true
@@ -36062,7 +36066,7 @@
         "max_tokens": 118000,
         "mode": "chat",
         "output_cost_per_token": 1.3e-07,
-        "source": "https://endpoints.ai.cloud.ovh.net/models/mistral-nemo-instruct-2407",
+        "source": "https://www.ovhcloud.com/en/public-cloud/ai-endpoints/catalog/",
         "supports_function_calling": true,
         "supports_response_schema": true,
         "supports_tool_choice": true
@@ -36075,7 +36079,7 @@
         "max_tokens": 128000,
         "mode": "chat",
         "output_cost_per_token": 2.8e-07,
-        "source": "https://endpoints.ai.cloud.ovh.net/models/mistral-small-3-2-24b-instruct-2506",
+        "source": "https://www.ovhcloud.com/en/public-cloud/ai-endpoints/catalog/",
         "supports_function_calling": true,
         "supports_response_schema": true,
         "supports_tool_choice": true,
@@ -36089,7 +36093,7 @@
         "max_tokens": 32000,
         "mode": "chat",
         "output_cost_per_token": 6.3e-07,
-        "source": "https://endpoints.ai.cloud.ovh.net/models/mixtral-8x7b-instruct-v0-1",
+        "source": "https://www.ovhcloud.com/en/public-cloud/ai-endpoints/catalog/",
         "supports_function_calling": false,
         "supports_response_schema": true,
         "supports_tool_choice": false
@@ -36102,7 +36106,7 @@
         "max_tokens": 32000,
         "mode": "chat",
         "output_cost_per_token": 8.7e-07,
-        "source": "https://endpoints.ai.cloud.ovh.net/models/qwen2-5-coder-32b-instruct",
+        "source": "https://www.ovhcloud.com/en/public-cloud/ai-endpoints/catalog/",
         "supports_function_calling": false,
         "supports_response_schema": true,
         "supports_tool_choice": false
@@ -36115,7 +36119,7 @@
         "max_tokens": 32000,
         "mode": "chat",
         "output_cost_per_token": 9.1e-07,
-        "source": "https://endpoints.ai.cloud.ovh.net/models/qwen2-5-vl-72b-instruct",
+        "source": "https://www.ovhcloud.com/en/public-cloud/ai-endpoints/catalog/",
         "supports_function_calling": false,
         "supports_response_schema": true,
         "supports_tool_choice": false,
@@ -36129,7 +36133,7 @@
         "max_tokens": 32000,
         "mode": "chat",
         "output_cost_per_token": 2.3e-07,
-        "source": "https://endpoints.ai.cloud.ovh.net/models/qwen3-32b",
+        "source": "https://www.ovhcloud.com/en/public-cloud/ai-endpoints/catalog/",
         "supports_function_calling": true,
         "supports_reasoning": true,
         "supports_response_schema": true,
@@ -36143,7 +36147,7 @@
         "max_tokens": 131000,
         "mode": "chat",
         "output_cost_per_token": 4e-07,
-        "source": "https://endpoints.ai.cloud.ovh.net/models/gpt-oss-120b",
+        "source": "https://www.ovhcloud.com/en/public-cloud/ai-endpoints/catalog/",
         "supports_function_calling": false,
         "supports_reasoning": true,
         "supports_response_schema": true,
@@ -36157,7 +36161,7 @@
         "max_tokens": 131000,
         "mode": "chat",
         "output_cost_per_token": 1.5e-07,
-        "source": "https://endpoints.ai.cloud.ovh.net/models/gpt-oss-20b",
+        "source": "https://www.ovhcloud.com/en/public-cloud/ai-endpoints/catalog/",
         "supports_function_calling": false,
         "supports_reasoning": true,
         "supports_response_schema": true,
@@ -36171,7 +36175,7 @@
         "max_tokens": 32000,
         "mode": "chat",
         "output_cost_per_token": 2.9e-07,
-        "source": "https://endpoints.ai.cloud.ovh.net/models/llava-next-mistral-7b",
+        "source": "https://www.ovhcloud.com/en/public-cloud/ai-endpoints/catalog/",
         "supports_function_calling": false,
         "supports_response_schema": true,
         "supports_tool_choice": false,
@@ -36185,7 +36189,7 @@
         "max_tokens": 256000,
         "mode": "chat",
         "output_cost_per_token": 1.9e-07,
-        "source": "https://endpoints.ai.cloud.ovh.net/models/mamba-codestral-7b-v0-1",
+        "source": "https://www.ovhcloud.com/en/public-cloud/ai-endpoints/catalog/",
         "supports_function_calling": false,
         "supports_response_schema": true,
         "supports_tool_choice": false
@@ -38321,7 +38325,7 @@
         "source": "https://docs.mistral.ai/capabilities/code_generation/"
     },
     "text-embedding-004": {
-        "deprecation_date": "2026-01-14",
+        "deprecation_date": "2027-04-01",
         "input_cost_per_character": 2.5e-08,
         "input_cost_per_token": 1e-07,
         "litellm_provider": "vertex_ai-embedding-models",
@@ -38946,7 +38950,7 @@
         "max_input_tokens": 262144,
         "mode": "chat",
         "output_cost_per_token": 3.6e-06,
-        "source": "https://www.together.ai/models/Qwen/Qwen3.5-397B-A17B",
+        "source": "https://www.together.ai/models/qwen3-5-397b-a17b",
         "supports_function_calling": true,
         "supports_parallel_function_calling": true,
         "supports_prompt_caching": true,
@@ -43183,7 +43187,7 @@
         "max_tokens": 2000000,
         "mode": "chat",
         "output_cost_per_token": 5e-07,
-        "source": "https://docs.x.ai/docs/models (Vertex AI Model Garden)",
+        "source": "https://docs.x.ai/developers/models",
         "supports_function_calling": true,
         "supports_response_schema": true,
         "supports_tool_choice": true,
@@ -43199,7 +43203,7 @@
         "max_tokens": 2000000,
         "mode": "chat",
         "output_cost_per_token": 5e-07,
-        "source": "https://docs.x.ai/docs/models (Vertex AI Model Garden)",
+        "source": "https://docs.x.ai/developers/models",
         "supports_function_calling": true,
         "supports_reasoning": true,
         "supports_response_schema": true,
@@ -43216,7 +43220,7 @@
         "max_tokens": 2000000,
         "mode": "chat",
         "output_cost_per_token": 6e-06,
-        "source": "https://docs.x.ai/docs/models (Vertex AI Model Garden)",
+        "source": "https://docs.x.ai/developers/models",
         "supports_function_calling": true,
         "supports_response_schema": true,
         "supports_tool_choice": true,
@@ -43232,7 +43236,7 @@
         "max_tokens": 2000000,
         "mode": "chat",
         "output_cost_per_token": 6e-06,
-        "source": "https://docs.x.ai/docs/models (Vertex AI Model Garden)",
+        "source": "https://docs.x.ai/developers/models",
         "supports_function_calling": true,
         "supports_reasoning": true,
         "supports_response_schema": true,
@@ -43352,7 +43356,7 @@
         "max_tokens": 1024,
         "mode": "video_generation",
         "output_cost_per_second": 0.4,
-        "source": "https://cloud.google.com/vertex-ai/generative-ai/docs/model-reference/veo",
+        "source": "https://docs.cloud.google.com/gemini-enterprise-agent-platform/models/veo/3-1-generate",
         "supported_modalities": [
             "text"
         ],
@@ -43366,7 +43370,7 @@
         "max_tokens": 1024,
         "mode": "video_generation",
         "output_cost_per_second": 0.15,
-        "source": "https://cloud.google.com/vertex-ai/generative-ai/docs/model-reference/veo",
+        "source": "https://docs.cloud.google.com/gemini-enterprise-agent-platform/models/veo/3-1-generate",
         "supported_modalities": [
             "text"
         ],
@@ -43381,7 +43385,7 @@
         "max_tokens": 1024,
         "mode": "video_generation",
         "output_cost_per_second": 0.4,
-        "source": "https://cloud.google.com/vertex-ai/generative-ai/docs/model-reference/veo",
+        "source": "https://docs.cloud.google.com/gemini-enterprise-agent-platform/models/veo/3-1-generate",
         "supported_modalities": [
             "text"
         ],
@@ -43396,7 +43400,7 @@
         "max_tokens": 1024,
         "mode": "video_generation",
         "output_cost_per_second": 0.15,
-        "source": "https://cloud.google.com/vertex-ai/generative-ai/docs/model-reference/veo",
+        "source": "https://docs.cloud.google.com/gemini-enterprise-agent-platform/models/veo/3-1-generate",
         "supported_modalities": [
             "text"
         ],
@@ -44100,7 +44104,8 @@
         "output_cost_per_second": 0.0001,
         "supported_endpoints": [
             "/v1/audio/transcriptions"
-        ]
+        ],
+        "deprecation_date": "2027-02-26"
     },
     "xai/grok-3": {
         "cache_read_input_token_cost": 2e-07,
@@ -45088,7 +45093,7 @@
         "litellm_provider": "azure",
         "mode": "video_generation",
         "output_cost_per_video_per_second": 0.1,
-        "source": "https://azure.microsoft.com/en-us/products/ai-services/video-generation",
+        "source": "https://ai.azure.com/catalog/models/sora-2",
         "supported_modalities": [
             "text"
         ],
@@ -45100,7 +45105,7 @@
         "litellm_provider": "azure",
         "mode": "video_generation",
         "output_cost_per_video_per_second": 0.3,
-        "source": "https://azure.microsoft.com/en-us/products/ai-services/video-generation",
+        "source": "https://ai.azure.com/catalog/models/sora-2-pro",
         "supported_modalities": [
             "text"
         ],
@@ -45112,7 +45117,7 @@
         "litellm_provider": "azure",
         "mode": "video_generation",
         "output_cost_per_video_per_second": 0.5,
-        "source": "https://azure.microsoft.com/en-us/products/ai-services/video-generation",
+        "source": "https://ai.azure.com/catalog/models/sora-2-pro",
         "supported_modalities": [
             "text"
         ],
@@ -49322,7 +49327,7 @@
         "input_cost_per_second": 0.0002833333333333333,
         "litellm_provider": "openai",
         "mode": "audio_transcription",
-        "source": "https://platform.openai.com/docs/models/gpt-realtime-whisper",
+        "source": "https://developers.openai.com/api/docs/models/gpt-realtime-whisper",
         "supported_endpoints": [
             "/v1/realtime",
             "/v1/realtime/transcription_sessions"
@@ -50574,7 +50579,7 @@
         "max_output_tokens": 128000,
         "max_tokens": 128000,
         "mode": "chat",
-        "source": "https://www.volcengine.com/docs/82379/1330310",
+        "source": "https://docs.volcengine.com/docs/82379/1330310",
         "supports_function_calling": true,
         "supports_reasoning": true,
         "supports_tool_choice": false,
@@ -50612,7 +50617,7 @@
         "max_output_tokens": 128000,
         "max_tokens": 128000,
         "mode": "chat",
-        "source": "https://www.volcengine.com/docs/82379/1330310",
+        "source": "https://docs.volcengine.com/docs/82379/1330310",
         "supports_function_calling": true,
         "supports_reasoning": true,
         "supports_tool_choice": false,
@@ -50650,7 +50655,7 @@
         "max_output_tokens": 128000,
         "max_tokens": 128000,
         "mode": "chat",
-        "source": "https://www.volcengine.com/docs/82379/1330310",
+        "source": "https://docs.volcengine.com/docs/82379/1330310",
         "supports_function_calling": true,
         "supports_reasoning": true,
         "supports_tool_choice": false,
@@ -50688,7 +50693,7 @@
         "max_output_tokens": 128000,
         "max_tokens": 128000,
         "mode": "chat",
-        "source": "https://www.volcengine.com/docs/82379/1330310",
+        "source": "https://docs.volcengine.com/docs/82379/1330310",
         "supports_function_calling": true,
         "supports_reasoning": true,
         "supports_tool_choice": false,
@@ -51437,7 +51442,7 @@
         "supports_function_calling": true,
         "supports_assistant_prefill": true,
         "supports_reasoning": true,
-        "source": "https://pinstripes.io/pricing"
+        "source": "https://pinstripes.io/"
     },
     "pinstripes/ps/qwen3.6-35b-a3b": {
         "max_tokens": 131072,
@@ -51450,7 +51455,7 @@
         "supports_function_calling": true,
         "supports_assistant_prefill": true,
         "supports_reasoning": true,
-        "source": "https://pinstripes.io/pricing"
+        "source": "https://pinstripes.io/"
     },
     "pinstripes/ps/qwen3-30b-a3b": {
         "max_tokens": 131072,
@@ -51463,7 +51468,7 @@
         "supports_function_calling": true,
         "supports_assistant_prefill": true,
         "supports_reasoning": true,
-        "source": "https://pinstripes.io/pricing"
+        "source": "https://pinstripes.io/"
     },
     "pinstripes/ps/qwen3-coder-30b-a3b": {
         "max_tokens": 131072,
@@ -51476,7 +51481,7 @@
         "supports_function_calling": true,
         "supports_assistant_prefill": true,
         "supports_reasoning": false,
-        "source": "https://pinstripes.io/pricing"
+        "source": "https://pinstripes.io/"
     },
     "pinstripes/ps/deepseek-v4-flash": {
         "max_tokens": 163840,
@@ -51489,7 +51494,7 @@
         "supports_function_calling": true,
         "supports_assistant_prefill": true,
         "supports_reasoning": true,
-        "source": "https://pinstripes.io/pricing"
+        "source": "https://pinstripes.io/"
     },
     "pinstripes/ps/minimax-m2.7": {
         "max_tokens": 1000192,
@@ -51502,7 +51507,7 @@
         "supports_function_calling": true,
         "supports_assistant_prefill": true,
         "supports_reasoning": false,
-        "source": "https://pinstripes.io/pricing"
+        "source": "https://pinstripes.io/"
     },
     "darkbloom/gemma-4-26b": {
         "input_cost_per_token": 3e-08,
@@ -51606,7 +51611,7 @@
         "input_cost_per_second": 7.5e-05,
         "litellm_provider": "openai",
         "mode": "audio_transcription",
-        "source": "https://platform.openai.com/docs/models/gpt-transcribe",
+        "source": "https://developers.openai.com/api/docs/models/gpt-transcribe",
         "supported_endpoints": [
             "/v1/audio/transcriptions",
             "/v1/realtime/transcription_sessions"
@@ -51624,7 +51629,7 @@
         "input_cost_per_second": 0.0002833333333333333,
         "litellm_provider": "openai",
         "mode": "audio_transcription",
-        "source": "https://platform.openai.com/docs/models/gpt-live-transcribe",
+        "source": "https://developers.openai.com/api/docs/models/gpt-live-transcribe",
         "supported_endpoints": [
             "/v1/realtime",
             "/v1/realtime/transcription_sessions"
@@ -51645,7 +51650,7 @@
         "max_output_tokens": 2000,
         "max_tokens": 2000,
         "mode": "realtime",
-        "source": "https://platform.openai.com/docs/models/gpt-realtime-translate",
+        "source": "https://developers.openai.com/api/docs/models/gpt-realtime-translate",
         "supported_modalities": [
             "audio"
         ],
@@ -51673,7 +51678,7 @@
             "search_context_size_low": 0.01,
             "search_context_size_medium": 0.01
         },
-        "source": "https://docs.claude.com/en/docs/about-claude/models/overview",
+        "source": "https://platform.claude.com/docs/en/about-claude/models/overview",
         "supports_adaptive_thinking": true,
         "thinking_always_on": true,
         "supports_mid_conversation_system": true,
@@ -51712,7 +51717,7 @@
             "search_context_size_low": 0.01,
             "search_context_size_medium": 0.01
         },
-        "source": "https://docs.claude.com/en/docs/about-claude/models/overview",
+        "source": "https://platform.claude.com/docs/en/about-claude/models/overview",
         "supports_adaptive_thinking": true,
         "thinking_always_on": true,
         "supports_assistant_prefill": false,
diff --git a/litellm/proxy/management_endpoints/team_callback_endpoints.py b/litellm/proxy/management_endpoints/team_callback_endpoints.py
index 08346983f32..c2f5dbb4032 100644
--- a/litellm/proxy/management_endpoints/team_callback_endpoints.py
+++ b/litellm/proxy/management_endpoints/team_callback_endpoints.py
@@ -252,7 +252,7 @@ async def add_team_callbacks(
     Use this if if you want different teams to have different success/failure callbacks
 
     Parameters:
-    - callback_name (Literal["langfuse", "langsmith", "gcs"], required): The name of the callback to add
+    - callback_name (str, required): The name of the callback to add, e.g. "langfuse", "langsmith", "gcs", "newrelic". The value is validated against the callbacks that support team-scoped credentials
     - callback_type (Literal["success", "failure", "success_and_failure"], required): The type of callback to add. One of:
         - "success": Callback for successful LLM calls
         - "failure": Callback for failed LLM calls
@@ -268,6 +268,8 @@ async def add_team_callbacks(
         - langsmith_api_key: The API key for the Langsmith callback
         - langsmith_project: The project for the Langsmith callback
         - langsmith_base_url: The base URL for the Langsmith callback
+        - newrelic_api_key: The ingest license key for the team's New Relic account; routes both LLM/agent traces and cost metrics to that account. Requires the proxy to run with LITELLM_OTEL_V2=true, otherwise this callback is rejected with a 400
+        - newrelic_region: The New Relic region for the team's account ("us" or "eu"), riding the team's own key
 
     Example curl:
     ```
diff --git a/model_prices_and_context_window.json b/model_prices_and_context_window.json
index 05c1cfd3179..7071eaa0807 100644
--- a/model_prices_and_context_window.json
+++ b/model_prices_and_context_window.json
@@ -3726,7 +3726,7 @@
         "output_cost_per_token": 0,
         "litellm_provider": "azure_ai",
         "mode": "chat",
-        "source": "https://azure.microsoft.com/en-us/pricing/details/ai-services/",
+        "source": "https://azure.microsoft.com/en-us/pricing/details/ai-foundry-models/aoai/",
         "comment": "Flat cost of $0.14 per M input tokens for Azure AI Foundry Model Router infrastructure. Use pattern: azure_ai/model_router/<deployment-name> where deployment-name is your Azure deployment (e.g., azure-model-router)"
     },
     "azure/eu/gpt-4o-2024-08-06": {
@@ -5328,7 +5328,7 @@
         "input_cost_per_second": 0.0002833333333333333,
         "litellm_provider": "azure",
         "mode": "audio_transcription",
-        "source": "https://learn.microsoft.com/en-us/azure/ai-foundry/openai/concepts/gpt-realtime-whisper",
+        "source": "https://learn.microsoft.com/en-us/azure/foundry/openai/concepts/gpt-realtime-whisper",
         "supported_endpoints": [
             "/v1/realtime",
             "/v1/realtime/transcription_sessions"
@@ -9107,7 +9107,7 @@
         "mode": "embedding",
         "output_cost_per_token": 0.0,
         "output_vector_size": 1024,
-        "source": "https://azuremarketplace.microsoft.com/en-us/marketplace/apps/cohere.cohere-embed-v3-english-offer?tab=PlansAndPrice",
+        "source": "https://marketplace.microsoft.com/en-us/marketplace/apps/cohere.cohere-embed-v3-english-offer?tab=PlansAndPrice",
         "supports_embedding_image_input": true
     },
     "azure_ai/Cohere-embed-v3-multilingual": {
@@ -9118,7 +9118,7 @@
         "mode": "embedding",
         "output_cost_per_token": 0.0,
         "output_vector_size": 1024,
-        "source": "https://azuremarketplace.microsoft.com/en-us/marketplace/apps/cohere.cohere-embed-v3-english-offer?tab=PlansAndPrice",
+        "source": "https://marketplace.microsoft.com/en-us/marketplace/apps/cohere.cohere-embed-v3-english-offer?tab=PlansAndPrice",
         "supports_embedding_image_input": true
     },
     "azure_ai/FLUX-1.1-pro": {
@@ -9134,7 +9134,7 @@
         "litellm_provider": "azure_ai",
         "mode": "image_generation",
         "output_cost_per_image": 0.04,
-        "source": "https://azuremarketplace.microsoft.com/pt-br/marketplace/apps/cohere.cohere-embed-4-offer?tab=PlansAndPrice",
+        "source": "https://marketplace.microsoft.com/pt-br/marketplace/apps/cohere.cohere-embed-4-offer?tab=PlansAndPrice",
         "supported_endpoints": [
             "/v1/images/generations"
         ]
@@ -9467,7 +9467,7 @@
         "max_tokens": 2048,
         "mode": "chat",
         "output_cost_per_token": 3.7e-07,
-        "source": "https://azuremarketplace.microsoft.com/en/marketplace/apps/metagenai.meta-llama-3-2-11b-vision-instruct-offer?tab=Overview",
+        "source": "https://marketplace.microsoft.com/en/marketplace/apps/metagenai.meta-llama-3-2-11b-vision-instruct-offer?tab=Overview",
         "supports_function_calling": true,
         "supports_tool_choice": true,
         "supports_vision": true
@@ -9481,7 +9481,7 @@
         "max_tokens": 2048,
         "mode": "chat",
         "output_cost_per_token": 2.04e-06,
-        "source": "https://azuremarketplace.microsoft.com/en/marketplace/apps/metagenai.meta-llama-3-2-90b-vision-instruct-offer?tab=Overview",
+        "source": "https://marketplace.microsoft.com/en/marketplace/apps/metagenai.meta-llama-3-2-90b-vision-instruct-offer?tab=Overview",
         "supports_function_calling": true,
         "supports_tool_choice": true,
         "supports_vision": true
@@ -9494,7 +9494,7 @@
         "max_tokens": 2048,
         "mode": "chat",
         "output_cost_per_token": 7.1e-07,
-        "source": "https://azuremarketplace.microsoft.com/en/marketplace/apps/metagenai.llama-3-3-70b-instruct-offer?tab=Overview",
+        "source": "https://marketplace.microsoft.com/en/marketplace/apps/metagenai.llama-3-3-70b-instruct-offer?tab=Overview",
         "supports_function_calling": true,
         "supports_tool_choice": true
     },
@@ -9543,7 +9543,7 @@
         "max_tokens": 2048,
         "mode": "chat",
         "output_cost_per_token": 1.6e-05,
-        "source": "https://azuremarketplace.microsoft.com/en-us/marketplace/apps/metagenai.meta-llama-3-1-405b-instruct-offer?tab=PlansAndPrice",
+        "source": "https://marketplace.microsoft.com/en-us/marketplace/apps/metagenai.meta-llama-3-1-405b-instruct-offer?tab=PlansAndPrice",
         "supports_tool_choice": true
     },
     "azure_ai/Meta-Llama-3.1-70B-Instruct": {
@@ -9554,7 +9554,7 @@
         "max_tokens": 2048,
         "mode": "chat",
         "output_cost_per_token": 3.54e-06,
-        "source": "https://azuremarketplace.microsoft.com/en-us/marketplace/apps/metagenai.meta-llama-3-1-70b-instruct-offer?tab=PlansAndPrice",
+        "source": "https://marketplace.microsoft.com/en-us/marketplace/apps/metagenai.meta-llama-3-1-70b-instruct-offer?tab=PlansAndPrice",
         "supports_tool_choice": true
     },
     "azure_ai/Meta-Llama-3.1-8B-Instruct": {
@@ -9566,7 +9566,7 @@
         "max_tokens": 2048,
         "mode": "chat",
         "output_cost_per_token": 6.1e-07,
-        "source": "https://azuremarketplace.microsoft.com/en-us/marketplace/apps/metagenai.meta-llama-3-1-8b-instruct-offer?tab=PlansAndPrice",
+        "source": "https://marketplace.microsoft.com/en-us/marketplace/apps/metagenai.meta-llama-3-1-8b-instruct-offer?tab=PlansAndPrice",
         "supports_tool_choice": true
     },
     "azure_ai/Phi-3-medium-128k-instruct": {
@@ -9756,7 +9756,7 @@
         "supported_endpoints": [
             "/v1/ocr"
         ],
-        "source": "https://azure.microsoft.com/en-us/pricing/details/ai-foundry-models/"
+        "source": "https://ai.azure.com/catalog/models/mistral-document-ai-2512"
     },
     "azure_ai/doc-intelligence/prebuilt-read": {
         "litellm_provider": "azure_ai",
@@ -9967,7 +9967,7 @@
         "mode": "embedding",
         "output_cost_per_token": 0.0,
         "output_vector_size": 3072,
-        "source": "https://azuremarketplace.microsoft.com/pt-br/marketplace/apps/cohere.cohere-embed-4-offer?tab=PlansAndPrice",
+        "source": "https://marketplace.microsoft.com/pt-br/marketplace/apps/cohere.cohere-embed-4-offer?tab=PlansAndPrice",
         "supported_endpoints": [
             "/v1/embeddings"
         ],
@@ -10151,7 +10151,7 @@
         "max_tokens": 8192,
         "mode": "chat",
         "output_cost_per_token": 0.00971,
-        "source": "https://azure.microsoft.com/en-us/products/ai-services/ai-foundry/models/jais-30b-chat"
+        "source": "https://ai.azure.com/catalog/models/jais-30b-chat"
     },
     "azure_ai/jamba-instruct": {
         "input_cost_per_token": 5e-07,
@@ -10208,7 +10208,7 @@
         "max_tokens": 4096,
         "mode": "chat",
         "output_cost_per_token": 4e-08,
-        "source": "https://azuremarketplace.microsoft.com/en/marketplace/apps/000-000.ministral-3b-2410-offer?tab=Overview",
+        "source": "https://marketplace.microsoft.com/en/marketplace/apps/000-000.ministral-3b-2410-offer?tab=Overview",
         "supports_function_calling": true,
         "supports_tool_choice": true
     },
@@ -10231,7 +10231,7 @@
         "max_tokens": 4096,
         "mode": "chat",
         "output_cost_per_token": 6e-06,
-        "source": "https://azuremarketplace.microsoft.com/en/marketplace/apps/000-000.mistral-ai-large-2407-offer?tab=Overview",
+        "source": "https://marketplace.microsoft.com/en/marketplace/apps/000-000.mistral-ai-large-2407-offer?tab=Overview",
         "supports_function_calling": true,
         "supports_tool_choice": true
     },
@@ -10243,7 +10243,7 @@
         "max_tokens": 4096,
         "mode": "chat",
         "output_cost_per_token": 6e-06,
-        "source": "https://azuremarketplace.microsoft.com/en/marketplace/apps/000-000.mistral-ai-large-2407-offer?tab=Overview",
+        "source": "https://marketplace.microsoft.com/en/marketplace/apps/000-000.mistral-ai-large-2407-offer?tab=Overview",
         "supports_function_calling": true,
         "supports_tool_choice": true
     },
@@ -10280,7 +10280,7 @@
         "max_tokens": 4096,
         "mode": "chat",
         "output_cost_per_token": 1.5e-07,
-        "source": "https://azuremarketplace.microsoft.com/en/marketplace/apps/000-000.mistral-nemo-12b-2407?tab=PlansAndPrice",
+        "source": "https://marketplace.microsoft.com/en/marketplace/apps/000-000.mistral-nemo-12b-2407?tab=PlansAndPrice",
         "supports_function_calling": true
     },
     "azure_ai/mistral-small": {
@@ -12315,7 +12315,8 @@
         "output_cost_per_token": 1e-05,
         "supported_endpoints": [
             "/v1/audio/transcriptions"
-        ]
+        ],
+        "deprecation_date": "2027-02-26"
     },
     "claude-haiku-4-5-20251001": {
         "deprecation_date": "2026-10-15",
@@ -17767,7 +17768,8 @@
         "supports_response_schema": true,
         "supports_tool_choice": true,
         "supports_vision": true,
-        "prompt_cache_min_tokens": 1024
+        "prompt_cache_min_tokens": 1024,
+        "deprecation_date": "2027-01-08"
     },
     "eu.anthropic.claude-opus-4-20250514-v1:0": {
         "cache_creation_input_token_cost": 1.875e-05,
@@ -22843,7 +22845,7 @@
         "output_cost_per_reasoning_token": 3e-06,
         "output_cost_per_token": 3e-06,
         "rpm": 2000,
-        "source": "https://ai.google.dev/pricing/gemini-3",
+        "source": "https://ai.google.dev/gemini-api/docs/pricing",
         "supported_endpoints": [
             "/v1/chat/completions",
             "/v1/completions",
@@ -22897,7 +22899,7 @@
         "output_cost_per_reasoning_token": 9e-06,
         "output_cost_per_token": 9e-06,
         "rpm": 2000,
-        "source": "https://ai.google.dev/pricing/gemini-3",
+        "source": "https://ai.google.dev/gemini-api/docs/pricing",
         "supported_endpoints": [
             "/v1/chat/completions",
             "/v1/completions",
@@ -22962,7 +22964,7 @@
         "output_cost_per_token_batches": 1.875e-06,
         "output_cost_per_token_flex": 1.875e-06,
         "rpm": 2000,
-        "source": "https://ai.google.dev/pricing/gemini-3",
+        "source": "https://ai.google.dev/gemini-api/docs/pricing",
         "supported_endpoints": [
             "/v1/chat/completions",
             "/v1/completions",
@@ -23021,7 +23023,7 @@
         "output_cost_per_token_batches": 1.875e-06,
         "output_cost_per_token_flex": 1.875e-06,
         "rpm": 2000,
-        "source": "https://ai.google.dev/pricing/gemini-3",
+        "source": "https://ai.google.dev/gemini-api/docs/pricing",
         "supported_endpoints": [
             "/v1/chat/completions",
             "/v1/completions",
@@ -23226,7 +23228,7 @@
         "mode": "chat",
         "output_cost_per_reasoning_token": 3e-06,
         "output_cost_per_token": 3e-06,
-        "source": "https://ai.google.dev/pricing/gemini-3",
+        "source": "https://ai.google.dev/gemini-api/docs/pricing",
         "supported_endpoints": [
             "/v1/chat/completions",
             "/v1/completions",
@@ -23310,7 +23312,7 @@
         "mode": "chat",
         "output_cost_per_reasoning_token": 9e-06,
         "output_cost_per_token": 9e-06,
-        "source": "https://ai.google.dev/pricing/gemini-3",
+        "source": "https://ai.google.dev/gemini-api/docs/pricing",
         "supported_endpoints": [
             "/v1/chat/completions",
             "/v1/completions",
@@ -23373,7 +23375,7 @@
         "output_cost_per_token": 3.75e-06,
         "output_cost_per_token_batches": 1.875e-06,
         "output_cost_per_token_flex": 1.875e-06,
-        "source": "https://ai.google.dev/pricing/gemini-3",
+        "source": "https://ai.google.dev/gemini-api/docs/pricing",
         "supported_endpoints": [
             "/v1/chat/completions",
             "/v1/completions",
@@ -23430,7 +23432,7 @@
         "output_cost_per_token": 3.75e-06,
         "output_cost_per_token_batches": 1.875e-06,
         "output_cost_per_token_flex": 1.875e-06,
-        "source": "https://ai.google.dev/pricing/gemini-3",
+        "source": "https://ai.google.dev/gemini-api/docs/pricing",
         "supported_endpoints": [
             "/v1/chat/completions",
             "/v1/completions",
@@ -25327,7 +25329,7 @@
         "supports_vision": true
     },
     "gpt-4o-audio-preview": {
-        "deprecation_date": "2027-01-20",
+        "deprecation_date": "2026-05-07",
         "input_cost_per_audio_token": 4e-05,
         "input_cost_per_token": 2.5e-06,
         "litellm_provider": "openai",
@@ -25650,7 +25652,7 @@
         "supports_vision": true
     },
     "gpt-4o-mini-audio-preview": {
-        "deprecation_date": "2027-01-20",
+        "deprecation_date": "2026-05-07",
         "input_cost_per_audio_token": 1e-05,
         "input_cost_per_token": 1.5e-07,
         "litellm_provider": "openai",
@@ -25688,7 +25690,7 @@
     "gpt-4o-mini-realtime-preview": {
         "cache_creation_input_audio_token_cost": 3e-07,
         "cache_read_input_token_cost": 3e-07,
-        "deprecation_date": "2027-01-20",
+        "deprecation_date": "2026-05-07",
         "input_cost_per_audio_token": 1e-05,
         "input_cost_per_token": 6e-07,
         "litellm_provider": "openai",
@@ -25787,7 +25789,8 @@
         "output_cost_per_token": 5e-06,
         "supported_endpoints": [
             "/v1/audio/transcriptions"
-        ]
+        ],
+        "deprecation_date": "2027-02-26"
     },
     "gpt-4o-mini-tts": {
         "input_cost_per_token": 2.5e-06,
@@ -25809,7 +25812,7 @@
     },
     "gpt-4o-realtime-preview": {
         "cache_read_input_token_cost": 2.5e-06,
-        "deprecation_date": "2027-01-20",
+        "deprecation_date": "2026-05-07",
         "input_cost_per_audio_token": 4e-05,
         "input_cost_per_token": 5e-06,
         "litellm_provider": "openai",
@@ -25828,7 +25831,7 @@
     },
     "gpt-4o-realtime-preview-2024-12-17": {
         "cache_read_input_token_cost": 2.5e-06,
-        "deprecation_date": "2027-01-20",
+        "deprecation_date": "2026-05-07",
         "input_cost_per_audio_token": 4e-05,
         "input_cost_per_token": 5e-06,
         "litellm_provider": "openai",
@@ -25847,7 +25850,7 @@
     },
     "gpt-4o-realtime-preview-2025-06-03": {
         "cache_read_input_token_cost": 2.5e-06,
-        "deprecation_date": "2027-01-20",
+        "deprecation_date": "2026-05-07",
         "input_cost_per_audio_token": 4e-05,
         "input_cost_per_token": 5e-06,
         "litellm_provider": "openai",
@@ -25926,7 +25929,8 @@
         "output_cost_per_token": 1e-05,
         "supported_endpoints": [
             "/v1/audio/transcriptions"
-        ]
+        ],
+        "deprecation_date": "2027-02-26"
     },
     "gpt-image-1.5": {
         "cache_read_input_token_cost": 1.25e-06,
@@ -27077,7 +27081,7 @@
         "supports_tool_choice": true,
         "supports_vision": true,
         "supports_web_search": true,
-        "source": "https://platform.openai.com/docs/models/gpt-5.6-cyber",
+        "source": "https://developers.openai.com/api/docs/models/gpt-5.6-cyber",
         "supports_computer_use": true,
         "supports_parallel_function_calling": true
     },
@@ -27116,7 +27120,7 @@
         "supports_tool_choice": true,
         "supports_vision": true,
         "supports_web_search": true,
-        "source": "https://platform.openai.com/docs/models/daybreak-red-latest",
+        "source": "https://developers.openai.com/api/docs/models/daybreak-red-latest",
         "supports_computer_use": true,
         "supports_parallel_function_calling": true
     },
@@ -27156,7 +27160,7 @@
         "supports_tool_choice": true,
         "supports_vision": true,
         "supports_web_search": true,
-        "source": "https://platform.openai.com/docs/models/daybreak-blue-latest",
+        "source": "https://developers.openai.com/api/docs/models/daybreak-blue-latest",
         "supports_parallel_function_calling": true
     },
     "chat-latest": {
@@ -27168,7 +27172,7 @@
         "max_tokens": 128000,
         "mode": "chat",
         "output_cost_per_token": 3e-05,
-        "source": "https://platform.openai.com/docs/models/chat-latest",
+        "source": "https://developers.openai.com/api/docs/models/chat-latest",
         "supported_endpoints": [
             "/v1/chat/completions",
             "/v1/responses"
@@ -30463,7 +30467,7 @@
             "search_context_size_low": 0.0025,
             "search_context_size_medium": 0.0025
         },
-        "source": "https://dev.meta.ai/docs/getting-started/pricing-rate-limits",
+        "source": "https://ai.developer.meta.com/docs/pricing-rate-limits",
         "supported_endpoints": [
             "/v1/chat/completions",
             "/v1/responses",
@@ -30504,7 +30508,7 @@
             "search_context_size_low": 0.0025,
             "search_context_size_medium": 0.0025
         },
-        "source": "https://dev.meta.ai/docs/getting-started/pricing-rate-limits",
+        "source": "https://ai.developer.meta.com/docs/pricing-rate-limits",
         "supported_endpoints": [
             "/v1/chat/completions",
             "/v1/responses",
@@ -30545,7 +30549,7 @@
             "search_context_size_low": 0.0025,
             "search_context_size_medium": 0.0025
         },
-        "source": "https://dev.meta.ai/docs/getting-started/pricing-rate-limits",
+        "source": "https://ai.developer.meta.com/docs/pricing-rate-limits",
         "supported_endpoints": [
             "/v1/chat/completions",
             "/v1/responses",
@@ -30578,7 +30582,7 @@
         "max_output_tokens": 4028,
         "max_tokens": 4028,
         "mode": "chat",
-        "source": "https://llama.developer.meta.com/docs/models",
+        "source": "https://ai.developer.meta.com/docs/models",
         "supported_modalities": [
             "text"
         ],
@@ -30594,7 +30598,7 @@
         "max_output_tokens": 4028,
         "max_tokens": 4028,
         "mode": "chat",
-        "source": "https://llama.developer.meta.com/docs/models",
+        "source": "https://ai.developer.meta.com/docs/models",
         "supported_modalities": [
             "text"
         ],
@@ -30610,7 +30614,7 @@
         "max_output_tokens": 4028,
         "max_tokens": 4028,
         "mode": "chat",
-        "source": "https://llama.developer.meta.com/docs/models",
+        "source": "https://ai.developer.meta.com/docs/models",
         "supported_modalities": [
             "text",
             "image"
@@ -30627,7 +30631,7 @@
         "max_output_tokens": 4028,
         "max_tokens": 4028,
         "mode": "chat",
-        "source": "https://llama.developer.meta.com/docs/models",
+        "source": "https://ai.developer.meta.com/docs/models",
         "supported_modalities": [
             "text",
             "image"
@@ -32572,7 +32576,7 @@
         "mode": "chat",
         "supports_function_calling": true,
         "supports_reasoning": true,
-        "source": "https://nebius.com/prices-ai-studio"
+        "source": "https://nebius.com/prices"
     },
     "nebius/deepseek-ai/DeepSeek-R1-0528": {
         "max_tokens": 164000,
@@ -32584,7 +32588,7 @@
         "mode": "chat",
         "supports_function_calling": true,
         "supports_reasoning": true,
-        "source": "https://nebius.com/prices-ai-studio"
+        "source": "https://nebius.com/prices"
     },
     "nebius/deepseek-ai/DeepSeek-R1-Distill-Llama-70B": {
         "max_tokens": 128000,
@@ -32595,7 +32599,7 @@
         "litellm_provider": "nebius",
         "mode": "chat",
         "supports_function_calling": true,
-        "source": "https://nebius.com/prices-ai-studio"
+        "source": "https://nebius.com/prices"
     },
     "nebius/deepseek-ai/DeepSeek-V3": {
         "max_tokens": 128000,
@@ -32606,7 +32610,7 @@
         "litellm_provider": "nebius",
         "mode": "chat",
         "supports_function_calling": true,
-        "source": "https://nebius.com/prices-ai-studio"
+        "source": "https://nebius.com/prices"
     },
     "nebius/deepseek-ai/DeepSeek-V3-0324": {
         "max_tokens": 128000,
@@ -32617,7 +32621,7 @@
         "litellm_provider": "nebius",
         "mode": "chat",
         "supports_function_calling": true,
-        "source": "https://nebius.com/prices-ai-studio"
+        "source": "https://nebius.com/prices"
     },
     "nebius/google/gemma-3-27b-it": {
         "max_tokens": 128000,
@@ -32629,7 +32633,7 @@
         "mode": "chat",
         "supports_function_calling": true,
         "supports_vision": true,
-        "source": "https://nebius.com/prices-ai-studio"
+        "source": "https://nebius.com/prices"
     },
     "nebius/meta-llama/Llama-3.3-70B-Instruct": {
         "max_tokens": 128000,
@@ -32640,7 +32644,7 @@
         "litellm_provider": "nebius",
         "mode": "chat",
         "supports_function_calling": true,
-        "source": "https://nebius.com/prices-ai-studio"
+        "source": "https://nebius.com/prices"
     },
     "nebius/meta-llama/Llama-Guard-3-8B": {
         "max_tokens": 128000,
@@ -32650,7 +32654,7 @@
         "output_cost_per_token": 6e-08,
         "litellm_provider": "nebius",
         "mode": "chat",
-        "source": "https://nebius.com/prices-ai-studio"
+        "source": "https://nebius.com/prices"
     },
     "nebius/meta-llama/Meta-Llama-3.1-8B-Instruct": {
         "max_tokens": 128000,
@@ -32661,7 +32665,7 @@
         "litellm_provider": "nebius",
         "mode": "chat",
         "supports_function_calling": true,
-        "source": "https://nebius.com/prices-ai-studio"
+        "source": "https://nebius.com/prices"
     },
     "nebius/meta-llama/Meta-Llama-3.1-70B-Instruct": {
         "max_tokens": 128000,
@@ -32672,7 +32676,7 @@
         "litellm_provider": "nebius",
         "mode": "chat",
         "supports_function_calling": true,
-        "source": "https://nebius.com/prices-ai-studio"
+        "source": "https://nebius.com/prices"
     },
     "nebius/meta-llama/Meta-Llama-3.1-405B-Instruct": {
         "max_tokens": 128000,
@@ -32683,7 +32687,7 @@
         "litellm_provider": "nebius",
         "mode": "chat",
         "supports_function_calling": true,
-        "source": "https://nebius.com/prices-ai-studio"
+        "source": "https://nebius.com/prices"
     },
     "nebius/mistralai/Mistral-Nemo-Instruct-2407": {
         "max_tokens": 128000,
@@ -32694,7 +32698,7 @@
         "litellm_provider": "nebius",
         "mode": "chat",
         "supports_function_calling": true,
-        "source": "https://nebius.com/prices-ai-studio"
+        "source": "https://nebius.com/prices"
     },
     "nebius/NousResearch/Hermes-3-Llama-3.1-405B": {
         "max_tokens": 128000,
@@ -32705,7 +32709,7 @@
         "litellm_provider": "nebius",
         "mode": "chat",
         "supports_function_calling": true,
-        "source": "https://nebius.com/prices-ai-studio"
+        "source": "https://nebius.com/prices"
     },
     "nebius/nvidia/Llama-3.1-Nemotron-Ultra-253B-v1": {
         "max_tokens": 128000,
@@ -32716,7 +32720,7 @@
         "litellm_provider": "nebius",
         "mode": "chat",
         "supports_function_calling": true,
-        "source": "https://nebius.com/prices-ai-studio"
+        "source": "https://nebius.com/prices"
     },
     "nebius/nvidia/Llama-3.3-Nemotron-Super-49B-v1": {
         "max_tokens": 131072,
@@ -32727,7 +32731,7 @@
         "litellm_provider": "nebius",
         "mode": "chat",
         "supports_function_calling": true,
-        "source": "https://nebius.com/prices-ai-studio"
+        "source": "https://nebius.com/prices"
     },
     "nebius/Qwen/Qwen3-235B-A22B": {
         "max_tokens": 262144,
@@ -32738,7 +32742,7 @@
         "litellm_provider": "nebius",
         "mode": "chat",
         "supports_function_calling": true,
-        "source": "https://nebius.com/prices-ai-studio"
+        "source": "https://nebius.com/prices"
     },
     "nebius/Qwen/Qwen3-32B": {
         "max_tokens": 32768,
@@ -32749,7 +32753,7 @@
         "litellm_provider": "nebius",
         "mode": "chat",
         "supports_function_calling": true,
-        "source": "https://nebius.com/prices-ai-studio"
+        "source": "https://nebius.com/prices"
     },
     "nebius/Qwen/Qwen3-30B-A3B": {
         "max_tokens": 32768,
@@ -32760,7 +32764,7 @@
         "litellm_provider": "nebius",
         "mode": "chat",
         "supports_function_calling": true,
-        "source": "https://nebius.com/prices-ai-studio"
+        "source": "https://nebius.com/prices"
     },
     "nebius/Qwen/Qwen3-14B": {
         "max_tokens": 32768,
@@ -32771,7 +32775,7 @@
         "litellm_provider": "nebius",
         "mode": "chat",
         "supports_function_calling": true,
-        "source": "https://nebius.com/prices-ai-studio"
+        "source": "https://nebius.com/prices"
     },
     "nebius/Qwen/Qwen3-4B": {
         "max_tokens": 32768,
@@ -32782,7 +32786,7 @@
         "litellm_provider": "nebius",
         "mode": "chat",
         "supports_function_calling": true,
-        "source": "https://nebius.com/prices-ai-studio"
+        "source": "https://nebius.com/prices"
     },
     "nebius/Qwen/QwQ-32B": {
         "max_tokens": 32768,
@@ -32794,7 +32798,7 @@
         "mode": "chat",
         "supports_function_calling": true,
         "supports_reasoning": true,
-        "source": "https://nebius.com/prices-ai-studio"
+        "source": "https://nebius.com/prices"
     },
     "nebius/Qwen/Qwen2.5-72B-Instruct": {
         "max_tokens": 128000,
@@ -32805,7 +32809,7 @@
         "litellm_provider": "nebius",
         "mode": "chat",
         "supports_function_calling": true,
-        "source": "https://nebius.com/prices-ai-studio"
+        "source": "https://nebius.com/prices"
     },
     "nebius/Qwen/Qwen2.5-32B-Instruct": {
         "max_tokens": 128000,
@@ -32816,7 +32820,7 @@
         "litellm_provider": "nebius",
         "mode": "chat",
         "supports_function_calling": true,
-        "source": "https://nebius.com/prices-ai-studio"
+        "source": "https://nebius.com/prices"
     },
     "nebius/Qwen/Qwen2.5-Coder-7B": {
         "max_tokens": 32768,
@@ -32827,7 +32831,7 @@
         "litellm_provider": "nebius",
         "mode": "chat",
         "supports_function_calling": true,
-        "source": "https://nebius.com/prices-ai-studio"
+        "source": "https://nebius.com/prices"
     },
     "nebius/Qwen/Qwen2.5-VL-72B-Instruct": {
         "max_tokens": 131072,
@@ -32839,7 +32843,7 @@
         "mode": "chat",
         "supports_function_calling": true,
         "supports_vision": true,
-        "source": "https://nebius.com/prices-ai-studio"
+        "source": "https://nebius.com/prices"
     },
     "nebius/Qwen/Qwen2-VL-72B-Instruct": {
         "max_tokens": 131072,
@@ -32851,7 +32855,7 @@
         "mode": "chat",
         "supports_function_calling": true,
         "supports_vision": true,
-        "source": "https://nebius.com/prices-ai-studio"
+        "source": "https://nebius.com/prices"
     },
     "nebius/Qwen/Qwen2-VL-7B-Instruct": {
         "max_tokens": 131072,
@@ -32862,7 +32866,7 @@
         "litellm_provider": "nebius",
         "mode": "chat",
         "supports_vision": true,
-        "source": "https://nebius.com/prices-ai-studio"
+        "source": "https://nebius.com/prices"
     },
     "nebius/BAAI/bge-en-icl": {
         "max_tokens": 32768,
@@ -32871,7 +32875,7 @@
         "output_cost_per_token": 0.0,
         "litellm_provider": "nebius",
         "mode": "embedding",
-        "source": "https://nebius.com/prices-ai-studio"
+        "source": "https://nebius.com/prices"
     },
     "nebius/BAAI/bge-multilingual-gemma2": {
         "max_tokens": 8192,
@@ -32880,7 +32884,7 @@
         "output_cost_per_token": 0.0,
         "litellm_provider": "nebius",
         "mode": "embedding",
-        "source": "https://nebius.com/prices-ai-studio"
+        "source": "https://nebius.com/prices"
     },
     "nebius/intfloat/e5-mistral-7b-instruct": {
         "max_tokens": 32768,
@@ -32889,7 +32893,7 @@
         "output_cost_per_token": 0.0,
         "litellm_provider": "nebius",
         "mode": "embedding",
-        "source": "https://nebius.com/prices-ai-studio"
+        "source": "https://nebius.com/prices"
     },
     "nvidia.nemotron-nano-12b-v2": {
         "input_cost_per_token": 2e-07,
@@ -33630,7 +33634,7 @@
         "max_tokens": 4000,
         "mode": "chat",
         "output_cost_per_token": 1.56e-06,
-        "source": "https://www.oracle.com/cloud/ai/generative-ai/pricing/",
+        "source": "https://www.oracle.com/artificial-intelligence/enterprise-ai/cost-estimator/",
         "supports_function_calling": true,
         "supports_response_schema": false,
         "supports_native_streaming": true
@@ -33643,7 +33647,7 @@
         "max_tokens": 4000,
         "mode": "chat",
         "output_cost_per_token": 1.56e-06,
-        "source": "https://www.oracle.com/cloud/ai/generative-ai/pricing/",
+        "source": "https://www.oracle.com/artificial-intelligence/enterprise-ai/cost-estimator/",
         "supports_function_calling": true,
         "supports_response_schema": false,
         "supports_native_streaming": true
@@ -33656,7 +33660,7 @@
         "max_tokens": 4000,
         "mode": "chat",
         "output_cost_per_token": 1.56e-06,
-        "source": "https://www.oracle.com/cloud/ai/generative-ai/pricing/",
+        "source": "https://www.oracle.com/artificial-intelligence/enterprise-ai/cost-estimator/",
         "supports_function_calling": true,
         "supports_response_schema": false,
         "supports_native_streaming": true
@@ -33713,7 +33717,7 @@
         "max_tokens": 8192,
         "mode": "chat",
         "output_cost_per_token": 1.56e-06,
-        "source": "https://www.oracle.com/cloud/ai/generative-ai/pricing/",
+        "source": "https://www.oracle.com/artificial-intelligence/enterprise-ai/cost-estimator/",
         "supports_function_calling": true,
         "supports_response_schema": false,
         "supports_native_streaming": true,
@@ -33727,7 +33731,7 @@
         "max_tokens": 8192,
         "mode": "chat",
         "output_cost_per_token": 1.56e-06,
-        "source": "https://www.oracle.com/cloud/ai/generative-ai/pricing/",
+        "source": "https://www.oracle.com/artificial-intelligence/enterprise-ai/cost-estimator/",
         "supports_function_calling": false,
         "supports_response_schema": false,
         "supports_native_streaming": true
@@ -33738,7 +33742,7 @@
         "max_input_tokens": 512,
         "mode": "embedding",
         "output_vector_size": 1024,
-        "source": "https://www.oracle.com/cloud/ai/generative-ai/pricing/",
+        "source": "https://www.oracle.com/artificial-intelligence/enterprise-ai/cost-estimator/",
         "supports_vision": true
     },
     "oci/cohere.command-a-reasoning-08-2025": {
@@ -34635,7 +34639,7 @@
         "max_tokens": 2048,
         "mode": "chat",
         "output_cost_per_token": 2e-07,
-        "source": "https://openrouter.ai/api/v1/models/bytedance/ui-tars-1.5-7b",
+        "source": "https://openrouter.ai/bytedance/ui-tars-1.5-7b",
         "supports_tool_choice": true
     },
     "openrouter/deepseek/deepseek-chat": {
@@ -34870,7 +34874,7 @@
         "output_cost_per_reasoning_token": 3e-06,
         "output_cost_per_token": 3e-06,
         "rpm": 2000,
-        "source": "https://ai.google.dev/pricing/gemini-3",
+        "source": "https://ai.google.dev/gemini-api/docs/pricing",
         "supported_endpoints": [
             "/v1/chat/completions",
             "/v1/completions",
@@ -34911,7 +34915,7 @@
         "output_cost_per_reasoning_token": 1.5e-06,
         "output_cost_per_token": 1.5e-06,
         "rpm": 2000,
-        "source": "https://ai.google.dev/pricing/gemini-3",
+        "source": "https://ai.google.dev/gemini-api/docs/pricing",
         "supported_endpoints": [
             "/v1/chat/completions",
             "/v1/completions",
@@ -35996,7 +36000,7 @@
         "max_tokens": 131000,
         "mode": "chat",
         "output_cost_per_token": 6.7e-07,
-        "source": "https://endpoints.ai.cloud.ovh.net/models/deepseek-r1-distill-llama-70b",
+        "source": "https://www.ovhcloud.com/en/public-cloud/ai-endpoints/catalog/",
         "supports_function_calling": true,
         "supports_reasoning": true,
         "supports_response_schema": true,
@@ -36010,7 +36014,7 @@
         "max_tokens": 131000,
         "mode": "chat",
         "output_cost_per_token": 1e-07,
-        "source": "https://endpoints.ai.cloud.ovh.net/models/llama-3-1-8b-instruct",
+        "source": "https://www.ovhcloud.com/en/public-cloud/ai-endpoints/catalog/",
         "supports_function_calling": true,
         "supports_response_schema": true,
         "supports_tool_choice": true
@@ -36023,7 +36027,7 @@
         "max_tokens": 131000,
         "mode": "chat",
         "output_cost_per_token": 6.7e-07,
-        "source": "https://endpoints.ai.cloud.ovh.net/models/meta-llama-3-1-70b-instruct",
+        "source": "https://www.ovhcloud.com/en/public-cloud/ai-endpoints/catalog/",
         "supports_function_calling": false,
         "supports_response_schema": false,
         "supports_tool_choice": false
@@ -36036,7 +36040,7 @@
         "max_tokens": 131000,
         "mode": "chat",
         "output_cost_per_token": 6.7e-07,
-        "source": "https://endpoints.ai.cloud.ovh.net/models/meta-llama-3-3-70b-instruct",
+        "source": "https://www.ovhcloud.com/en/public-cloud/ai-endpoints/catalog/",
         "supports_function_calling": true,
         "supports_response_schema": true,
         "supports_tool_choice": true
@@ -36049,7 +36053,7 @@
         "max_tokens": 127000,
         "mode": "chat",
         "output_cost_per_token": 1e-07,
-        "source": "https://endpoints.ai.cloud.ovh.net/models/mistral-7b-instruct-v0-3",
+        "source": "https://www.ovhcloud.com/en/public-cloud/ai-endpoints/catalog/",
         "supports_function_calling": true,
         "supports_response_schema": true,
         "supports_tool_choice": true
@@ -36062,7 +36066,7 @@
         "max_tokens": 118000,
         "mode": "chat",
         "output_cost_per_token": 1.3e-07,
-        "source": "https://endpoints.ai.cloud.ovh.net/models/mistral-nemo-instruct-2407",
+        "source": "https://www.ovhcloud.com/en/public-cloud/ai-endpoints/catalog/",
         "supports_function_calling": true,
         "supports_response_schema": true,
         "supports_tool_choice": true
@@ -36075,7 +36079,7 @@
         "max_tokens": 128000,
         "mode": "chat",
         "output_cost_per_token": 2.8e-07,
-        "source": "https://endpoints.ai.cloud.ovh.net/models/mistral-small-3-2-24b-instruct-2506",
+        "source": "https://www.ovhcloud.com/en/public-cloud/ai-endpoints/catalog/",
         "supports_function_calling": true,
         "supports_response_schema": true,
         "supports_tool_choice": true,
@@ -36089,7 +36093,7 @@
         "max_tokens": 32000,
         "mode": "chat",
         "output_cost_per_token": 6.3e-07,
-        "source": "https://endpoints.ai.cloud.ovh.net/models/mixtral-8x7b-instruct-v0-1",
+        "source": "https://www.ovhcloud.com/en/public-cloud/ai-endpoints/catalog/",
         "supports_function_calling": false,
         "supports_response_schema": true,
         "supports_tool_choice": false
@@ -36102,7 +36106,7 @@
         "max_tokens": 32000,
         "mode": "chat",
         "output_cost_per_token": 8.7e-07,
-        "source": "https://endpoints.ai.cloud.ovh.net/models/qwen2-5-coder-32b-instruct",
+        "source": "https://www.ovhcloud.com/en/public-cloud/ai-endpoints/catalog/",
         "supports_function_calling": false,
         "supports_response_schema": true,
         "supports_tool_choice": false
@@ -36115,7 +36119,7 @@
         "max_tokens": 32000,
         "mode": "chat",
         "output_cost_per_token": 9.1e-07,
-        "source": "https://endpoints.ai.cloud.ovh.net/models/qwen2-5-vl-72b-instruct",
+        "source": "https://www.ovhcloud.com/en/public-cloud/ai-endpoints/catalog/",
         "supports_function_calling": false,
         "supports_response_schema": true,
         "supports_tool_choice": false,
@@ -36129,7 +36133,7 @@
         "max_tokens": 32000,
         "mode": "chat",
         "output_cost_per_token": 2.3e-07,
-        "source": "https://endpoints.ai.cloud.ovh.net/models/qwen3-32b",
+        "source": "https://www.ovhcloud.com/en/public-cloud/ai-endpoints/catalog/",
         "supports_function_calling": true,
         "supports_reasoning": true,
         "supports_response_schema": true,
@@ -36143,7 +36147,7 @@
         "max_tokens": 131000,
         "mode": "chat",
         "output_cost_per_token": 4e-07,
-        "source": "https://endpoints.ai.cloud.ovh.net/models/gpt-oss-120b",
+        "source": "https://www.ovhcloud.com/en/public-cloud/ai-endpoints/catalog/",
         "supports_function_calling": false,
         "supports_reasoning": true,
         "supports_response_schema": true,
@@ -36157,7 +36161,7 @@
         "max_tokens": 131000,
         "mode": "chat",
         "output_cost_per_token": 1.5e-07,
-        "source": "https://endpoints.ai.cloud.ovh.net/models/gpt-oss-20b",
+        "source": "https://www.ovhcloud.com/en/public-cloud/ai-endpoints/catalog/",
         "supports_function_calling": false,
         "supports_reasoning": true,
         "supports_response_schema": true,
@@ -36171,7 +36175,7 @@
         "max_tokens": 32000,
         "mode": "chat",
         "output_cost_per_token": 2.9e-07,
-        "source": "https://endpoints.ai.cloud.ovh.net/models/llava-next-mistral-7b",
+        "source": "https://www.ovhcloud.com/en/public-cloud/ai-endpoints/catalog/",
         "supports_function_calling": false,
         "supports_response_schema": true,
         "supports_tool_choice": false,
@@ -36185,7 +36189,7 @@
         "max_tokens": 256000,
         "mode": "chat",
         "output_cost_per_token": 1.9e-07,
-        "source": "https://endpoints.ai.cloud.ovh.net/models/mamba-codestral-7b-v0-1",
+        "source": "https://www.ovhcloud.com/en/public-cloud/ai-endpoints/catalog/",
         "supports_function_calling": false,
         "supports_response_schema": true,
         "supports_tool_choice": false
@@ -38321,7 +38325,7 @@
         "source": "https://docs.mistral.ai/capabilities/code_generation/"
     },
     "text-embedding-004": {
-        "deprecation_date": "2026-01-14",
+        "deprecation_date": "2027-04-01",
         "input_cost_per_character": 2.5e-08,
         "input_cost_per_token": 1e-07,
         "litellm_provider": "vertex_ai-embedding-models",
@@ -38946,7 +38950,7 @@
         "max_input_tokens": 262144,
         "mode": "chat",
         "output_cost_per_token": 3.6e-06,
-        "source": "https://www.together.ai/models/Qwen/Qwen3.5-397B-A17B",
+        "source": "https://www.together.ai/models/qwen3-5-397b-a17b",
         "supports_function_calling": true,
         "supports_parallel_function_calling": true,
         "supports_prompt_caching": true,
@@ -43183,7 +43187,7 @@
         "max_tokens": 2000000,
         "mode": "chat",
         "output_cost_per_token": 5e-07,
-        "source": "https://docs.x.ai/docs/models (Vertex AI Model Garden)",
+        "source": "https://docs.x.ai/developers/models",
         "supports_function_calling": true,
         "supports_response_schema": true,
         "supports_tool_choice": true,
@@ -43199,7 +43203,7 @@
         "max_tokens": 2000000,
         "mode": "chat",
         "output_cost_per_token": 5e-07,
-        "source": "https://docs.x.ai/docs/models (Vertex AI Model Garden)",
+        "source": "https://docs.x.ai/developers/models",
         "supports_function_calling": true,
         "supports_reasoning": true,
         "supports_response_schema": true,
@@ -43216,7 +43220,7 @@
         "max_tokens": 2000000,
         "mode": "chat",
         "output_cost_per_token": 6e-06,
-        "source": "https://docs.x.ai/docs/models (Vertex AI Model Garden)",
+        "source": "https://docs.x.ai/developers/models",
         "supports_function_calling": true,
         "supports_response_schema": true,
         "supports_tool_choice": true,
@@ -43232,7 +43236,7 @@
         "max_tokens": 2000000,
         "mode": "chat",
         "output_cost_per_token": 6e-06,
-        "source": "https://docs.x.ai/docs/models (Vertex AI Model Garden)",
+        "source": "https://docs.x.ai/developers/models",
         "supports_function_calling": true,
         "supports_reasoning": true,
         "supports_response_schema": true,
@@ -43352,7 +43356,7 @@
         "max_tokens": 1024,
         "mode": "video_generation",
         "output_cost_per_second": 0.4,
-        "source": "https://cloud.google.com/vertex-ai/generative-ai/docs/model-reference/veo",
+        "source": "https://docs.cloud.google.com/gemini-enterprise-agent-platform/models/veo/3-1-generate",
         "supported_modalities": [
             "text"
         ],
@@ -43366,7 +43370,7 @@
         "max_tokens": 1024,
         "mode": "video_generation",
         "output_cost_per_second": 0.15,
-        "source": "https://cloud.google.com/vertex-ai/generative-ai/docs/model-reference/veo",
+        "source": "https://docs.cloud.google.com/gemini-enterprise-agent-platform/models/veo/3-1-generate",
         "supported_modalities": [
             "text"
         ],
@@ -43381,7 +43385,7 @@
         "max_tokens": 1024,
         "mode": "video_generation",
         "output_cost_per_second": 0.4,
-        "source": "https://cloud.google.com/vertex-ai/generative-ai/docs/model-reference/veo",
+        "source": "https://docs.cloud.google.com/gemini-enterprise-agent-platform/models/veo/3-1-generate",
         "supported_modalities": [
             "text"
         ],
@@ -43396,7 +43400,7 @@
         "max_tokens": 1024,
         "mode": "video_generation",
         "output_cost_per_second": 0.15,
-        "source": "https://cloud.google.com/vertex-ai/generative-ai/docs/model-reference/veo",
+        "source": "https://docs.cloud.google.com/gemini-enterprise-agent-platform/models/veo/3-1-generate",
         "supported_modalities": [
             "text"
         ],
@@ -44100,7 +44104,8 @@
         "output_cost_per_second": 0.0001,
         "supported_endpoints": [
             "/v1/audio/transcriptions"
-        ]
+        ],
+        "deprecation_date": "2027-02-26"
     },
     "xai/grok-3": {
         "cache_read_input_token_cost": 2e-07,
@@ -45088,7 +45093,7 @@
         "litellm_provider": "azure",
         "mode": "video_generation",
         "output_cost_per_video_per_second": 0.1,
-        "source": "https://azure.microsoft.com/en-us/products/ai-services/video-generation",
+        "source": "https://ai.azure.com/catalog/models/sora-2",
         "supported_modalities": [
             "text"
         ],
@@ -45100,7 +45105,7 @@
         "litellm_provider": "azure",
         "mode": "video_generation",
         "output_cost_per_video_per_second": 0.3,
-        "source": "https://azure.microsoft.com/en-us/products/ai-services/video-generation",
+        "source": "https://ai.azure.com/catalog/models/sora-2-pro",
         "supported_modalities": [
             "text"
         ],
@@ -45112,7 +45117,7 @@
         "litellm_provider": "azure",
         "mode": "video_generation",
         "output_cost_per_video_per_second": 0.5,
-        "source": "https://azure.microsoft.com/en-us/products/ai-services/video-generation",
+        "source": "https://ai.azure.com/catalog/models/sora-2-pro",
         "supported_modalities": [
             "text"
         ],
@@ -49322,7 +49327,7 @@
         "input_cost_per_second": 0.0002833333333333333,
         "litellm_provider": "openai",
         "mode": "audio_transcription",
-        "source": "https://platform.openai.com/docs/models/gpt-realtime-whisper",
+        "source": "https://developers.openai.com/api/docs/models/gpt-realtime-whisper",
         "supported_endpoints": [
             "/v1/realtime",
             "/v1/realtime/transcription_sessions"
@@ -50574,7 +50579,7 @@
         "max_output_tokens": 128000,
         "max_tokens": 128000,
         "mode": "chat",
-        "source": "https://www.volcengine.com/docs/82379/1330310",
+        "source": "https://docs.volcengine.com/docs/82379/1330310",
         "supports_function_calling": true,
         "supports_reasoning": true,
         "supports_tool_choice": false,
@@ -50612,7 +50617,7 @@
         "max_output_tokens": 128000,
         "max_tokens": 128000,
         "mode": "chat",
-        "source": "https://www.volcengine.com/docs/82379/1330310",
+        "source": "https://docs.volcengine.com/docs/82379/1330310",
         "supports_function_calling": true,
         "supports_reasoning": true,
         "supports_tool_choice": false,
@@ -50650,7 +50655,7 @@
         "max_output_tokens": 128000,
         "max_tokens": 128000,
         "mode": "chat",
-        "source": "https://www.volcengine.com/docs/82379/1330310",
+        "source": "https://docs.volcengine.com/docs/82379/1330310",
         "supports_function_calling": true,
         "supports_reasoning": true,
         "supports_tool_choice": false,
@@ -50688,7 +50693,7 @@
         "max_output_tokens": 128000,
         "max_tokens": 128000,
         "mode": "chat",
-        "source": "https://www.volcengine.com/docs/82379/1330310",
+        "source": "https://docs.volcengine.com/docs/82379/1330310",
         "supports_function_calling": true,
         "supports_reasoning": true,
         "supports_tool_choice": false,
@@ -51437,7 +51442,7 @@
         "supports_function_calling": true,
         "supports_assistant_prefill": true,
         "supports_reasoning": true,
-        "source": "https://pinstripes.io/pricing"
+        "source": "https://pinstripes.io/"
     },
     "pinstripes/ps/qwen3.6-35b-a3b": {
         "max_tokens": 131072,
@@ -51450,7 +51455,7 @@
         "supports_function_calling": true,
         "supports_assistant_prefill": true,
         "supports_reasoning": true,
-        "source": "https://pinstripes.io/pricing"
+        "source": "https://pinstripes.io/"
     },
     "pinstripes/ps/qwen3-30b-a3b": {
         "max_tokens": 131072,
@@ -51463,7 +51468,7 @@
         "supports_function_calling": true,
         "supports_assistant_prefill": true,
         "supports_reasoning": true,
-        "source": "https://pinstripes.io/pricing"
+        "source": "https://pinstripes.io/"
     },
     "pinstripes/ps/qwen3-coder-30b-a3b": {
         "max_tokens": 131072,
@@ -51476,7 +51481,7 @@
         "supports_function_calling": true,
         "supports_assistant_prefill": true,
         "supports_reasoning": false,
-        "source": "https://pinstripes.io/pricing"
+        "source": "https://pinstripes.io/"
     },
     "pinstripes/ps/deepseek-v4-flash": {
         "max_tokens": 163840,
@@ -51489,7 +51494,7 @@
         "supports_function_calling": true,
         "supports_assistant_prefill": true,
         "supports_reasoning": true,
-        "source": "https://pinstripes.io/pricing"
+        "source": "https://pinstripes.io/"
     },
     "pinstripes/ps/minimax-m2.7": {
         "max_tokens": 1000192,
@@ -51502,7 +51507,7 @@
         "supports_function_calling": true,
         "supports_assistant_prefill": true,
         "supports_reasoning": false,
-        "source": "https://pinstripes.io/pricing"
+        "source": "https://pinstripes.io/"
     },
     "darkbloom/gemma-4-26b": {
         "input_cost_per_token": 3e-08,
@@ -51606,7 +51611,7 @@
         "input_cost_per_second": 7.5e-05,
         "litellm_provider": "openai",
         "mode": "audio_transcription",
-        "source": "https://platform.openai.com/docs/models/gpt-transcribe",
+        "source": "https://developers.openai.com/api/docs/models/gpt-transcribe",
         "supported_endpoints": [
             "/v1/audio/transcriptions",
             "/v1/realtime/transcription_sessions"
@@ -51624,7 +51629,7 @@
         "input_cost_per_second": 0.0002833333333333333,
         "litellm_provider": "openai",
         "mode": "audio_transcription",
-        "source": "https://platform.openai.com/docs/models/gpt-live-transcribe",
+        "source": "https://developers.openai.com/api/docs/models/gpt-live-transcribe",
         "supported_endpoints": [
             "/v1/realtime",
             "/v1/realtime/transcription_sessions"
@@ -51645,7 +51650,7 @@
         "max_output_tokens": 2000,
         "max_tokens": 2000,
         "mode": "realtime",
-        "source": "https://platform.openai.com/docs/models/gpt-realtime-translate",
+        "source": "https://developers.openai.com/api/docs/models/gpt-realtime-translate",
         "supported_modalities": [
             "audio"
         ],
@@ -51673,7 +51678,7 @@
             "search_context_size_low": 0.01,
             "search_context_size_medium": 0.01
         },
-        "source": "https://docs.claude.com/en/docs/about-claude/models/overview",
+        "source": "https://platform.claude.com/docs/en/about-claude/models/overview",
         "supports_adaptive_thinking": true,
         "thinking_always_on": true,
         "supports_mid_conversation_system": true,
@@ -51712,7 +51717,7 @@
             "search_context_size_low": 0.01,
             "search_context_size_medium": 0.01
         },
-        "source": "https://docs.claude.com/en/docs/about-claude/models/overview",
+        "source": "https://platform.claude.com/docs/en/about-claude/models/overview",
         "supports_adaptive_thinking": true,
         "thinking_always_on": true,
         "supports_assistant_prefill": false,
diff --git a/scripts/auto-close-duplicates.test.ts b/scripts/auto-close-duplicates.test.ts
new file mode 100644
index 00000000000..b49bf05cbc2
--- /dev/null
+++ b/scripts/auto-close-duplicates.test.ts
@@ -0,0 +1,348 @@
+import { describe, expect, test } from "bun:test";
+
+import {
+  CLOSED_MARKER,
+  REOPEN_COMMENT,
+  candidateNumbers,
+  duplicateTarget,
+  normalizeTitle,
+  pendingNotice,
+  readConfig,
+  reopenTarget,
+  sweepClosedIssue,
+  sweepIssue,
+  type Comment,
+  type GitHubApi,
+  type Issue,
+  type Reaction,
+  type SweepConfig,
+} from "./auto-close-duplicates";
+
+const NOW = new Date("2026-09-04T09:00:00Z");
+const DAY_MS = 24 * 60 * 60 * 1000;
+const daysAgo = (days: number): string => new Date(NOW.getTime() - days * DAY_MS).toISOString();
+
+const issue = (number: number, title: string, overrides: Partial<Issue> = {}): Issue => ({
+  number,
+  title,
+  state: "open",
+  user: { login: "reporter" },
+  ...overrides,
+});
+
+const notice = (candidates: readonly number[], createdAt: string, overrides: Partial<Comment> = {}): Comment => ({
+  id: 900,
+  body: `<!-- litellm:potential-duplicate candidates=${candidates.join(",")}, -->\n**Potential duplicate detected**`,
+  created_at: createdAt,
+  user: { type: "Bot", login: "github-actions[bot]" },
+  ...overrides,
+});
+
+const humanComment = (createdAt: string, body = "It is not the same thing", login = "reporter"): Comment => ({
+  id: 901,
+  body,
+  created_at: createdAt,
+  user: { type: "User", login },
+});
+
+const config: SweepConfig = { repo: "BerriAI/litellm", graceDays: 3, dryRun: false, now: NOW };
+
+describe("normalizeTitle", () => {
+  test("drops the template prefix, case, and punctuation", () => {
+    expect(normalizeTitle("[Bug]: Gemma 4-e4b fails on Vertex!")).toBe("gemma 4 e4b fails on vertex");
+    expect(normalizeTitle("[Feature]:   ")).toBe("");
+  });
+});
+
+describe("candidateNumbers", () => {
+  test("reads only the marker field, keeps older issues, sorted ascending and deduplicated", () => {
+    const body = "<!-- litellm:potential-duplicate candidates=40,10,30,10, -->\n- #1 - see #1 (100% similar)";
+    expect(candidateNumbers(body, 35)).toEqual([10, 30]);
+  });
+
+  test("returns nothing without the marker", () => {
+    expect(candidateNumbers("- #1 - looks like #1", 35)).toEqual([]);
+  });
+});
+
+describe("pendingNotice", () => {
+  test("waits out the grace period from the latest notice", () => {
+    const fresh = pendingNotice(issue(35, "t"), [notice([10], daysAgo(2.9))], config);
+    expect(fresh.kind).toBe("skip");
+    const aged = pendingNotice(issue(35, "t"), [notice([10], daysAgo(3.1))], config);
+    expect(aged.kind).toBe("pending");
+    const reposted = pendingNotice(
+      issue(35, "t"),
+      [notice([10], daysAgo(6)), notice([10], daysAgo(1), { id: 902 })],
+      config,
+    );
+    expect(reposted.kind).toBe("skip");
+  });
+
+  test("an objection posted before a re-posted notice still keeps the issue open", () => {
+    const verdict = pendingNotice(
+      issue(35, "t"),
+      [notice([10], daysAgo(10)), humanComment(daysAgo(7)), notice([10], daysAgo(4), { id: 902 })],
+      config,
+    );
+    expect(verdict).toEqual({ kind: "skip", reason: "someone replied after the notice" });
+  });
+
+  test("a zero-day grace period acts on the notice at once", () => {
+    const verdict = pendingNotice(issue(35, "t"), [notice([10], daysAgo(0.01))], { ...config, graceDays: 0 });
+    expect(verdict.kind).toBe("pending");
+  });
+
+  test("a human reply after the notice keeps the issue open, a bot reply does not", () => {
+    const human = pendingNotice(issue(35, "t"), [notice([10], daysAgo(5)), humanComment(daysAgo(4))], config);
+    expect(human).toEqual({ kind: "skip", reason: "someone replied after the notice" });
+    const bot = pendingNotice(
+      issue(35, "t"),
+      [notice([10], daysAgo(5)), { id: 903, body: "triage", created_at: daysAgo(4), user: { type: "Bot", login: "triage[bot]" } }],
+      config,
+    );
+    expect(bot.kind).toBe("pending");
+  });
+
+  test("a human quoting the marker is not a notice", () => {
+    const quoted = pendingNotice(issue(35, "t"), [notice([10], daysAgo(5), { user: { type: "User", login: "reporter" } })], config);
+    expect(quoted).toEqual({ kind: "skip", reason: "carries no duplicate notice" });
+  });
+
+  test("never closes an issue twice: a reopened issue is left alone", () => {
+    const reopened = pendingNotice(
+      issue(35, "t"),
+      [notice([10], daysAgo(9)), { id: 904, body: `Closed automatically\n\n${CLOSED_MARKER}`, created_at: daysAgo(5), user: { type: "Bot", login: "github-actions[bot]" } }],
+      config,
+    );
+    expect(reopened).toEqual({ kind: "skip", reason: "was reopened after an automatic close" });
+  });
+
+  test("skips pull requests and issues whose only candidates are newer", () => {
+    expect(pendingNotice(issue(35, "t", { pull_request: {} }), [notice([10], daysAgo(5))], config).kind).toBe("skip");
+    expect(pendingNotice(issue(35, "t"), [notice([40], daysAgo(5))], config)).toEqual({
+      kind: "skip",
+      reason: "no candidate is older than this issue",
+    });
+  });
+});
+
+describe("duplicateTarget", () => {
+  const reporter = issue(35, "[Bug]: Gemma 4-e4b fails on Vertex");
+
+  test("closes only against the earliest open issue with the identical normalized title", () => {
+    const verdict = duplicateTarget(
+      reporter,
+      [issue(10, "[Bug]: Gemma 4-e4n fails on Vertex"), issue(20, "[bug]: gemma 4-e4b fails on vertex"), issue(30, "[Bug]: Gemma 4-e4b fails on Vertex")],
+      [],
+    );
+    expect(verdict).toEqual({ kind: "close", duplicateOf: 20 });
+  });
+
+  test("a near miss in the title is not a duplicate", () => {
+    const verdict = duplicateTarget(reporter, [issue(10, "[Bug]: Gemma 4-e4n fails on Vertex")], []);
+    expect(verdict).toEqual({ kind: "skip", reason: "no older open issue has the identical title" });
+  });
+
+  test("bare template titles never match each other", () => {
+    const verdict = duplicateTarget(issue(35, "[Bug]: "), [issue(10, "[Bug]: ")], []);
+    expect(verdict.kind).toBe("skip");
+    expect(verdict.kind === "skip" && verdict.reason).toContain("too short");
+  });
+
+  test("a closed candidate or a pull request is never the target", () => {
+    expect(duplicateTarget(reporter, [issue(10, reporter.title, { state: "closed" })], []).kind).toBe("skip");
+    expect(duplicateTarget(reporter, [issue(10, reporter.title, { pull_request: {} })], []).kind).toBe("skip");
+  });
+
+  test("a thumbs down on the notice keeps the issue open", () => {
+    const verdict = duplicateTarget(reporter, [issue(10, reporter.title)], [{ content: "+1" }, { content: "-1" }]);
+    expect(verdict).toEqual({ kind: "skip", reason: "someone gave the notice a thumbs down" });
+  });
+});
+
+describe("sweepIssue", () => {
+  const reporter = issue(35, "[Bug]: Gemma 4-e4b fails on Vertex");
+  const original = issue(10, "[Bug]: Gemma 4-e4b fails on Vertex");
+
+  function fakeApi(
+    comments: readonly Comment[] = [notice([10], daysAgo(5))],
+    reactionsByNotice: Readonly<Record<number, readonly Reaction[]>> = {},
+  ): { readonly api: GitHubApi; readonly writes: readonly string[] } {
+    const writes: string[] = [];
+    const api: GitHubApi = {
+      request: async <T>(method: string, path: string, body?: object): Promise<T> => {
+        if (method !== "GET") {
+          writes.push(`${method} ${path} ${JSON.stringify(body)}`);
+          return {} as T;
+        }
+        if (path.startsWith("/repos/BerriAI/litellm/issues/35/comments")) {
+          return comments as T;
+        }
+        const reactionsPath = path.match(/^\/repos\/BerriAI\/litellm\/issues\/comments\/(\d+)\/reactions/);
+        if (reactionsPath) {
+          return (reactionsByNotice[Number(reactionsPath[1])] ?? []) as T;
+        }
+        if (path === "/repos/BerriAI/litellm/issues/10") {
+          return original as T;
+        }
+        throw new Error(`unexpected GET ${path}`);
+      },
+    };
+    return { api, writes };
+  }
+
+  test("a dry run reports the close and writes nothing", async () => {
+    const { api, writes } = fakeApi();
+    const verdict = await sweepIssue(api, { ...config, dryRun: true }, reporter);
+    expect(verdict).toEqual({ kind: "close", duplicateOf: 10 });
+    expect(writes).toEqual([]);
+  });
+
+  test("a thumbs down on an earlier notice still keeps the issue open", async () => {
+    const { api, writes } = fakeApi([notice([10], daysAgo(9)), notice([10], daysAgo(5), { id: 902 })], { 900: [{ content: "-1" }] });
+    const verdict = await sweepIssue(api, config, reporter);
+    expect(verdict).toEqual({ kind: "skip", reason: "someone gave the notice a thumbs down" });
+    expect(writes).toEqual([]);
+  });
+
+  test("a real run comments, labels, then closes with the duplicate reason", async () => {
+    const { api, writes } = fakeApi();
+    const verdict = await sweepIssue(api, config, reporter);
+    expect(verdict).toEqual({ kind: "close", duplicateOf: 10 });
+    expect(writes.map((write) => write.split(" ").slice(0, 2).join(" "))).toEqual([
+      "POST /repos/BerriAI/litellm/issues/35/comments",
+      "POST /repos/BerriAI/litellm/issues/35/labels",
+      "PATCH /repos/BerriAI/litellm/issues/35",
+    ]);
+    expect(writes[0]).toContain("duplicate of #10");
+    expect(writes[0]).toContain("unanswered for 3 days");
+    expect(writes[0]).toContain(CLOSED_MARKER);
+    expect(writes[1]).toContain('{"labels":["duplicate"]}');
+    expect(writes[2]).toContain('{"state":"closed","state_reason":"duplicate"}');
+  });
+});
+
+describe("reopenTarget", () => {
+  const closedByBot = (overrides: Partial<Issue> = {}): Issue =>
+    issue(35, "t", { state: "closed", closed_by: { type: "Bot" }, ...overrides });
+  const closeMarker = (createdAt: string): Comment => ({
+    id: 905,
+    body: `Closed automatically as a duplicate of #10.\n\n${CLOSED_MARKER}`,
+    created_at: createdAt,
+    user: { type: "Bot", login: "github-actions[bot]" },
+  });
+
+  test("a reporter reply after the automatic close reopens", () => {
+    const verdict = reopenTarget(closedByBot(), [closeMarker(daysAgo(2)), humanComment(daysAgo(1))]);
+    expect(verdict).toEqual({ kind: "reopen" });
+  });
+
+  test("an issue closed by a person stays closed", () => {
+    const verdict = reopenTarget(closedByBot({ closed_by: { type: "User" } }), [
+      closeMarker(daysAgo(2)),
+      humanComment(daysAgo(1)),
+    ]);
+    expect(verdict).toEqual({ kind: "skip", reason: "was closed by a person" });
+  });
+
+  test("without the automatic-close marker nothing reopens", () => {
+    const verdict = reopenTarget(closedByBot(), [humanComment(daysAgo(1))]);
+    expect(verdict).toEqual({ kind: "skip", reason: "carries no automatic-close marker" });
+  });
+
+  test("a maintainer reply alone does not reopen", () => {
+    const verdict = reopenTarget(closedByBot(), [
+      closeMarker(daysAgo(2)),
+      humanComment(daysAgo(1), "Confirmed duplicate", "maintainer"),
+    ]);
+    expect(verdict).toEqual({ kind: "skip", reason: "the reporter has not replied since the close" });
+  });
+
+  test("a reporter comment from before the close does not reopen", () => {
+    const verdict = reopenTarget(closedByBot(), [humanComment(daysAgo(3)), closeMarker(daysAgo(2))]);
+    expect(verdict).toEqual({ kind: "skip", reason: "the reporter has not replied since the close" });
+  });
+
+  test("a pull request never reopens", () => {
+    const verdict = reopenTarget(closedByBot({ pull_request: {} }), [closeMarker(daysAgo(2)), humanComment(daysAgo(1))]);
+    expect(verdict).toEqual({ kind: "skip", reason: "is a pull request" });
+  });
+});
+
+describe("sweepClosedIssue", () => {
+  function fakeApi(issueBody: Issue, comments: readonly Comment[]): { readonly api: GitHubApi; readonly writes: readonly string[] } {
+    const writes: string[] = [];
+    const api: GitHubApi = {
+      request: async <T>(method: string, path: string, body?: object): Promise<T> => {
+        if (method !== "GET") {
+          writes.push(`${method} ${path} ${JSON.stringify(body)}`);
+          return {} as T;
+        }
+        if (path.startsWith("/repos/BerriAI/litellm/issues/35/comments")) {
+          return comments as T;
+        }
+        if (path === "/repos/BerriAI/litellm/issues/35") {
+          return issueBody as T;
+        }
+        throw new Error(`unexpected GET ${path}`);
+      },
+    };
+    return { api, writes };
+  }
+
+  const closedByBot = issue(35, "t", { state: "closed", closed_by: { type: "Bot" } });
+  const closeMarker: Comment = {
+    id: 905,
+    body: `Closed automatically as a duplicate of #10.\n\n${CLOSED_MARKER}`,
+    created_at: daysAgo(2),
+    user: { type: "Bot", login: "github-actions[bot]" },
+  };
+
+  test("a real run unlabels, reopens, then explains", async () => {
+    const { api, writes } = fakeApi(closedByBot, [closeMarker, humanComment(daysAgo(1))]);
+    const verdict = await sweepClosedIssue(api, config, 35);
+    expect(verdict).toEqual({ kind: "reopen" });
+    expect(writes).toEqual([
+      "DELETE /repos/BerriAI/litellm/issues/35/labels/duplicate undefined",
+      'PATCH /repos/BerriAI/litellm/issues/35 {"state":"open"}',
+      `POST /repos/BerriAI/litellm/issues/35/comments {"body":"${REOPEN_COMMENT}"}`,
+    ]);
+  });
+
+  test("a dry run reports the reopen and writes nothing", async () => {
+    const { api, writes } = fakeApi(closedByBot, [closeMarker, humanComment(daysAgo(1))]);
+    const verdict = await sweepClosedIssue(api, { ...config, dryRun: true }, 35);
+    expect(verdict).toEqual({ kind: "reopen" });
+    expect(writes).toEqual([]);
+  });
+});
+
+describe("readConfig", () => {
+  test("defaults to a real run with a 3-day grace period", () => {
+    const parsed = readConfig({ GITHUB_TOKEN: "t", GITHUB_REPOSITORY: "BerriAI/litellm" }, NOW);
+    expect(parsed).toEqual({ token: "t", repo: "BerriAI/litellm", graceDays: 3, dryRun: false, now: NOW });
+  });
+
+  test("honors DRY_RUN and GRACE_PERIOD_DAYS overrides", () => {
+    const parsed = readConfig(
+      { GITHUB_TOKEN: "t", GITHUB_REPOSITORY: "o/r", DRY_RUN: "true", GRACE_PERIOD_DAYS: "0" },
+      NOW,
+    );
+    expect(parsed.dryRun).toBe(true);
+    expect(parsed.graceDays).toBe(0);
+  });
+
+  test("an empty GRACE_PERIOD_DAYS, as a schedule run renders it, means the default", () => {
+    const parsed = readConfig({ GITHUB_TOKEN: "t", GITHUB_REPOSITORY: "o/r", GRACE_PERIOD_DAYS: "" }, NOW);
+    expect(parsed.graceDays).toBe(3);
+  });
+
+  test("refuses a missing token, a malformed repository, or a bad grace period", () => {
+    expect(() => readConfig({ GITHUB_REPOSITORY: "o/r" }, NOW)).toThrow("GITHUB_TOKEN");
+    expect(() => readConfig({ GITHUB_TOKEN: "t", GITHUB_REPOSITORY: "litellm" }, NOW)).toThrow("owner/repo");
+    expect(() => readConfig({ GITHUB_TOKEN: "t", GITHUB_REPOSITORY: "o/r", GRACE_PERIOD_DAYS: "-1" }, NOW)).toThrow(
+      "GRACE_PERIOD_DAYS",
+    );
+  });
+});
diff --git a/scripts/auto-close-duplicates.ts b/scripts/auto-close-duplicates.ts
new file mode 100644
index 00000000000..c595104d886
--- /dev/null
+++ b/scripts/auto-close-duplicates.ts
@@ -0,0 +1,300 @@
+#!/usr/bin/env bun
+
+declare const process: { readonly env: Readonly<Record<string, string | undefined>> };
+
+export interface Issue {
+  readonly number: number;
+  readonly title: string;
+  readonly state: string;
+  readonly user: { readonly login: string };
+  readonly closed_by?: { readonly type: string } | null;
+  readonly pull_request?: unknown;
+}
+
+export interface Comment {
+  readonly id: number;
+  readonly body: string;
+  readonly created_at: string;
+  readonly user: { readonly type: string; readonly login: string };
+}
+
+export interface Reaction {
+  readonly content: string;
+}
+
+export interface GitHubApi {
+  readonly request: <T>(method: "GET" | "POST" | "PATCH" | "DELETE", path: string, body?: object) => Promise<T>;
+}
+
+export interface SweepConfig {
+  readonly repo: string;
+  readonly graceDays: number;
+  readonly dryRun: boolean;
+  readonly now: Date;
+}
+
+export type NoticeVerdict =
+  | { readonly kind: "pending"; readonly notices: readonly Comment[]; readonly candidates: readonly number[] }
+  | { readonly kind: "skip"; readonly reason: string };
+
+export type CloseVerdict =
+  | { readonly kind: "close"; readonly duplicateOf: number }
+  | { readonly kind: "skip"; readonly reason: string };
+
+export type ReopenVerdict =
+  | { readonly kind: "reopen" }
+  | { readonly kind: "skip"; readonly reason: string };
+
+export const FLAG_LABEL = "potential-duplicate";
+export const CLOSED_MARKER = "<!-- litellm:closed-as-duplicate -->";
+export const DEFAULT_GRACE_DAYS = 3;
+export const REOPEN_COMMENT =
+  "Reopened automatically: the reporter replied after the duplicate close, so this needs a human look.";
+const NOTICE_MARKER = /<!-- litellm:potential-duplicate candidates=([\d,]*) -->/;
+const MIN_TITLE_WORDS = 3;
+const PAGE_SIZE = 100;
+const DAY_MS = 24 * 60 * 60 * 1000;
+const REOPEN_LOOKBACK_DAYS = 30;
+
+const skip = (reason: string): { readonly kind: "skip"; readonly reason: string } => ({ kind: "skip", reason });
+
+export function normalizeTitle(title: string): string {
+  return title
+    .toLowerCase()
+    .replace(/^\s*\[[^\]]*\]\s*:?/, "")
+    .replace(/[^a-z0-9]+/g, " ")
+    .trim();
+}
+
+export function candidateNumbers(noticeBody: string, issueNumber: number): readonly number[] {
+  const field = noticeBody.match(NOTICE_MARKER);
+  if (!field) {
+    return [];
+  }
+  const older = field[1]
+    .split(",")
+    .filter((value) => value !== "")
+    .map(Number)
+    .filter((candidate) => candidate < issueNumber);
+  return [...new Set(older)].sort((a, b) => a - b);
+}
+
+export function pendingNotice(
+  issue: Issue,
+  comments: readonly Comment[],
+  config: Pick<SweepConfig, "graceDays" | "now">,
+): NoticeVerdict {
+  if (issue.pull_request !== undefined) {
+    return skip("is a pull request");
+  }
+  if (comments.some((comment) => comment.body.includes(CLOSED_MARKER))) {
+    return skip("was reopened after an automatic close");
+  }
+  const notices = comments.filter((comment) => comment.user.type === "Bot" && NOTICE_MARKER.test(comment.body));
+  const first = notices[0];
+  const latest = notices[notices.length - 1];
+  if (first === undefined || latest === undefined) {
+    return skip("carries no duplicate notice");
+  }
+  const ageDays = (config.now.getTime() - new Date(latest.created_at).getTime()) / DAY_MS;
+  if (ageDays < config.graceDays) {
+    return skip(`notice is ${ageDays.toFixed(1)} days old, grace period is ${config.graceDays}`);
+  }
+  const firstNoticeAt = new Date(first.created_at);
+  if (comments.some((comment) => comment.user.type !== "Bot" && new Date(comment.created_at) > firstNoticeAt)) {
+    return skip("someone replied after the notice");
+  }
+  const candidates = candidateNumbers(latest.body, issue.number);
+  if (candidates.length === 0) {
+    return skip("no candidate is older than this issue");
+  }
+  return { kind: "pending", notices, candidates };
+}
+
+export function duplicateTarget(
+  issue: Issue,
+  candidates: readonly Issue[],
+  reactions: readonly Reaction[],
+): CloseVerdict {
+  if (reactions.some((reaction) => reaction.content === "-1")) {
+    return skip("someone gave the notice a thumbs down");
+  }
+  const title = normalizeTitle(issue.title);
+  if (title.split(" ").length < MIN_TITLE_WORDS) {
+    return skip(`title "${issue.title}" is too short to match on`);
+  }
+  const original = candidates.find(
+    (candidate) =>
+      candidate.state === "open" && candidate.pull_request === undefined && normalizeTitle(candidate.title) === title,
+  );
+  if (original === undefined) {
+    return skip("no older open issue has the identical title");
+  }
+  return { kind: "close", duplicateOf: original.number };
+}
+
+export function reopenTarget(issue: Issue, comments: readonly Comment[]): ReopenVerdict {
+  if (issue.pull_request !== undefined) {
+    return skip("is a pull request");
+  }
+  if (issue.closed_by?.type !== "Bot") {
+    return skip("was closed by a person");
+  }
+  const marker = comments.find((comment) => comment.body.includes(CLOSED_MARKER));
+  if (marker === undefined) {
+    return skip("carries no automatic-close marker");
+  }
+  const markerAt = new Date(marker.created_at);
+  if (!comments.some((comment) => comment.user.login === issue.user.login && new Date(comment.created_at) > markerAt)) {
+    return skip("the reporter has not replied since the close");
+  }
+  return { kind: "reopen" };
+}
+
+export function closingComment(duplicateOf: number, graceDays: number): string {
+  return `Closed automatically as a duplicate of #${duplicateOf}. Its title is identical to that older open issue and the duplicate notice above went unanswered for ${graceDays} days. If this is wrong, comment here with how it differs from #${duplicateOf} and this issue will be reopened automatically within a day.
+
+${CLOSED_MARKER}`;
+}
+
+async function listAll<T>(api: GitHubApi, path: string, page = 1): Promise<readonly T[]> {
+  const separator = path.includes("?") ? "&" : "?";
+  const batch = await api.request<readonly T[]>("GET", `${path}${separator}per_page=${PAGE_SIZE}&page=${page}`);
+  return batch.length < PAGE_SIZE ? batch : [...batch, ...(await listAll<T>(api, path, page + 1))];
+}
+
+async function closeAsDuplicate(
+  api: GitHubApi,
+  config: SweepConfig,
+  issueNumber: number,
+  duplicateOf: number,
+): Promise<void> {
+  const issuePath = `/repos/${config.repo}/issues/${issueNumber}`;
+  await api.request("POST", `${issuePath}/comments`, { body: closingComment(duplicateOf, config.graceDays) });
+  await api.request("POST", `${issuePath}/labels`, { labels: ["duplicate"] });
+  await api.request("PATCH", issuePath, { state: "closed", state_reason: "duplicate" });
+}
+
+async function reopenForReporter(api: GitHubApi, config: SweepConfig, issueNumber: number): Promise<void> {
+  const issuePath = `/repos/${config.repo}/issues/${issueNumber}`;
+  await api.request("DELETE", `${issuePath}/labels/duplicate`);
+  await api.request("PATCH", issuePath, { state: "open" });
+  await api.request("POST", `${issuePath}/comments`, { body: REOPEN_COMMENT });
+}
+
+export async function sweepClosedIssue(api: GitHubApi, config: SweepConfig, issueNumber: number): Promise<ReopenVerdict> {
+  const issue = await api.request<Issue>("GET", `/repos/${config.repo}/issues/${issueNumber}`);
+  const comments = await listAll<Comment>(api, `/repos/${config.repo}/issues/${issueNumber}/comments`);
+  const verdict = reopenTarget(issue, comments);
+  if (verdict.kind === "reopen" && !config.dryRun) {
+    await reopenForReporter(api, config, issueNumber);
+  }
+  return verdict;
+}
+
+export async function sweepIssue(api: GitHubApi, config: SweepConfig, issue: Issue): Promise<CloseVerdict> {
+  const comments = await listAll<Comment>(api, `/repos/${config.repo}/issues/${issue.number}/comments`);
+  const pending = pendingNotice(issue, comments, config);
+  if (pending.kind === "skip") {
+    return pending;
+  }
+  const reactions = (
+    await Promise.all(
+      pending.notices.map((notice) => listAll<Reaction>(api, `/repos/${config.repo}/issues/comments/${notice.id}/reactions`)),
+    )
+  ).flat();
+  const candidates = await Promise.all(
+    pending.candidates.map((candidate) => api.request<Issue>("GET", `/repos/${config.repo}/issues/${candidate}`)),
+  );
+  const verdict = duplicateTarget(issue, candidates, reactions);
+  if (verdict.kind === "close" && !config.dryRun) {
+    await closeAsDuplicate(api, config, issue.number, verdict.duplicateOf);
+  }
+  return verdict;
+}
+
+function describe(issue: Issue, verdict: CloseVerdict, dryRun: boolean): string {
+  if (verdict.kind === "skip") {
+    return `#${issue.number}: skipped, ${verdict.reason}`;
+  }
+  return `#${issue.number}: ${dryRun ? "would close" : "closed"} as a duplicate of #${verdict.duplicateOf}`;
+}
+
+export async function sweep(api: GitHubApi, config: SweepConfig): Promise<readonly CloseVerdict[]> {
+  const issues = await listAll<Issue>(api, `/repos/${config.repo}/issues?state=open&labels=${FLAG_LABEL}`);
+  console.log(`${issues.length} open issues carry the ${FLAG_LABEL} label in ${config.repo}${config.dryRun ? " (dry run)" : ""}`);
+  return issues.reduce<Promise<readonly CloseVerdict[]>>(async (previous, issue) => {
+    const verdicts = await previous;
+    const verdict = await sweepIssue(api, config, issue);
+    console.log(describe(issue, verdict, config.dryRun));
+    return [...verdicts, verdict];
+  }, Promise.resolve([]));
+}
+
+function describeReopen(issueNumber: number, verdict: ReopenVerdict, dryRun: boolean): string {
+  if (verdict.kind === "skip") {
+    return `#${issueNumber}: skipped, ${verdict.reason}`;
+  }
+  return `#${issueNumber}: ${dryRun ? "would reopen" : "reopened"} for the reporter's reply`;
+}
+
+export async function reopenSweep(api: GitHubApi, config: SweepConfig): Promise<readonly ReopenVerdict[]> {
+  const since = new Date(config.now.getTime() - REOPEN_LOOKBACK_DAYS * DAY_MS).toISOString();
+  const closedPath = `/repos/${config.repo}/issues?state=closed&labels=duplicate,${FLAG_LABEL}&since=${encodeURIComponent(since)}`;
+  const issues = await listAll<Issue>(api, closedPath);
+  console.log(`${issues.length} recently closed issues carry the duplicate and ${FLAG_LABEL} labels in ${config.repo}${config.dryRun ? " (dry run)" : ""}`);
+  return issues.reduce<Promise<readonly ReopenVerdict[]>>(async (previous, issue) => {
+    const verdicts = await previous;
+    const verdict = await sweepClosedIssue(api, config, issue.number);
+    console.log(describeReopen(issue.number, verdict, config.dryRun));
+    return [...verdicts, verdict];
+  }, Promise.resolve([]));
+}
+
+export function readConfig(env: Readonly<Record<string, string | undefined>>, now: Date): SweepConfig & { readonly token: string } {
+  const token = env.GITHUB_TOKEN;
+  const repo = env.GITHUB_REPOSITORY;
+  if (!token || !repo || !/^[\w.-]+\/[\w.-]+$/.test(repo)) {
+    throw new Error("GITHUB_TOKEN and GITHUB_REPOSITORY (owner/repo) are required");
+  }
+  const rawGraceDays = env.GRACE_PERIOD_DAYS?.trim();
+  const graceDays = rawGraceDays === undefined || rawGraceDays === "" ? DEFAULT_GRACE_DAYS : Number(rawGraceDays);
+  if (!Number.isFinite(graceDays) || graceDays < 0) {
+    throw new Error(`GRACE_PERIOD_DAYS must be a non-negative number, got "${env.GRACE_PERIOD_DAYS}"`);
+  }
+  return { token, repo, graceDays, dryRun: env.DRY_RUN === "true", now };
+}
+
+export function githubApi(token: string): GitHubApi {
+  return {
+    request: async <T>(method: "GET" | "POST" | "PATCH" | "DELETE", path: string, body?: object): Promise<T> => {
+      const response = await fetch(`https://api.github.com${path}`, {
+        method,
+        headers: {
+          Authorization: `Bearer ${token}`,
+          Accept: "application/vnd.github+json",
+          "X-GitHub-Api-Version": "2022-11-28",
+          "User-Agent": "litellm-auto-close-duplicates",
+          ...(body === undefined ? {} : { "Content-Type": "application/json" }),
+        },
+        body: body === undefined ? undefined : JSON.stringify(body),
+      });
+      if (!response.ok) {
+        throw new Error(`${method} ${path} failed: ${response.status} ${response.statusText}`);
+      }
+      return (await response.json()) as T;
+    },
+  };
+}
+
+if (import.meta.main) {
+  const { token, ...config } = readConfig(process.env, new Date());
+  const api = githubApi(token);
+  const closeVerdicts = await sweep(api, config);
+  const reopenVerdicts = await reopenSweep(api, config);
+  const closed = closeVerdicts.filter((verdict) => verdict.kind === "close").length;
+  const reopened = reopenVerdicts.filter((verdict) => verdict.kind === "reopen").length;
+  console.log(
+    `${config.dryRun ? "Would close" : "Closed"} ${closed} of ${closeVerdicts.length} flagged issues, ${config.dryRun ? "would reopen" : "reopened"} ${reopened}`,
+  );
+}
diff --git a/tests/proxy_unit_tests/test_proxy_utils.py b/tests/proxy_unit_tests/test_proxy_utils.py
index 3bde72ccd49..35de9961054 100644
--- a/tests/proxy_unit_tests/test_proxy_utils.py
+++ b/tests/proxy_unit_tests/test_proxy_utils.py
@@ -1317,6 +1317,62 @@ def test_proxy_config_state_post_init_callback_call(monkeypatch):
     assert config["litellm_settings"]["default_team_settings"][0]["team_id"] == "test"
 
 
+@pytest.mark.asyncio
+async def test_default_team_settings_newrelic_resolves_traces_and_metrics():
+    """Static `default_team_settings` is the config-file twin of POST /team/callback.
+
+    A team pinned to New Relic through `default_team_settings` must reach the
+    same two loggers the dynamic path does: the per-team metrics logger (cost
+    and usage) and the trace logger (LLM/agent spans). This proves the static
+    path resolves both, not just one, so the config-file customer gets the
+    same per-team routing as the API customer.
+    """
+    from litellm.litellm_core_utils.litellm_logging import Logging
+    from litellm.proxy.litellm_pre_call_utils import LiteLLMProxyRequestSetup
+    from litellm.proxy.proxy_server import ProxyConfig
+
+    pc = ProxyConfig()
+    pc.config = {
+        "litellm_settings": {
+            "default_team_settings": [
+                {
+                    "team_id": "team-a",
+                    "success_callback": ["newrelic"],
+                    "newrelic_api_key": "team-a-ingest-key",
+                    "newrelic_region": "eu",
+                }
+            ]
+        }
+    }
+
+    callback_metadata = LiteLLMProxyRequestSetup.add_team_based_callbacks_from_config(
+        team_id="team-a",
+        proxy_config=pc,
+    )
+
+    assert callback_metadata is not None
+    assert callback_metadata.success_callback == ["newrelic"]
+    assert callback_metadata.callback_vars == {
+        "newrelic_api_key": "team-a-ingest-key",
+        "newrelic_region": "eu",
+    }
+
+    logging_obj = Logging(
+        model="gpt-3.5-turbo",
+        messages=[{"role": "user", "content": "hi"}],
+        stream=False,
+        call_type="completion",
+        start_time=None,
+        litellm_call_id="static-nr-1",
+        function_id="static-nr-1",
+    )
+    logging_obj._trusted_callback_vars = tuple(callback_metadata.callback_vars.items())
+
+    resolved = logging_obj._resolve_dynamic_callback_string("newrelic")
+    resolved_names = {type(logger).__name__ for logger in resolved}
+    assert resolved_names == {"NewRelicMetricsLogger", "NewRelicLogger"}
+
+
 def test_proxy_config_state_get_config_state_error():
     """
     Ensures that get_config_state does not raise an error when the config is not a valid dictionary
diff --git a/ui/litellm-dashboard/src/lib/http/schema.d.ts b/ui/litellm-dashboard/src/lib/http/schema.d.ts
index 8a564a07489..81a2fefe32e 100644
--- a/ui/litellm-dashboard/src/lib/http/schema.d.ts
+++ b/ui/litellm-dashboard/src/lib/http/schema.d.ts
@@ -15505,7 +15505,7 @@ export interface paths {
          *     Use this if if you want different teams to have different success/failure callbacks
          *
          *     Parameters:
-         *     - callback_name (Literal["langfuse", "langsmith", "gcs"], required): The name of the callback to add
+         *     - callback_name (str, required): The name of the callback to add, e.g. "langfuse", "langsmith", "gcs", "newrelic". The value is validated against the callbacks that support team-scoped credentials
          *     - callback_type (Literal["success", "failure", "success_and_failure"], required): The type of callback to add. One of:
          *         - "success": Callback for successful LLM calls
          *         - "failure": Callback for failed LLM calls
@@ -15521,6 +15521,8 @@ export interface paths {
          *         - langsmith_api_key: The API key for the Langsmith callback
          *         - langsmith_project: The project for the Langsmith callback
          *         - langsmith_base_url: The base URL for the Langsmith callback
+         *         - newrelic_api_key: The ingest license key for the team's New Relic account; routes both LLM/agent traces and cost metrics to that account. Requires the proxy to run with LITELLM_OTEL_V2=true, otherwise this callback is rejected with a 400
+         *         - newrelic_region: The New Relic region for the team's account ("us" or "eu"), riding the team's own key
          *
          *     Example curl:
          *     ```