From c0de87d08d4bee0069280071a0fb1854bdd0a646 Mon Sep 17 00:00:00 2001
From: mateo-berri <277851410+mateo-berri@users.noreply.github.com>
Date: Thu, 30 Jul 2026 17:46:37 -0700
Subject: [PATCH] fix(proxy): resolve team-alias models in the stream usage
 support gate

Team-scoped models store an internal model_name_{team_id}_{uuid} name
with the public alias only in team_public_model_name, so resolving them
through get_model_list without team_id returned no deployments and the
gate skipped injection, leaving those streams on tiktoken estimates.
Thread user_api_key_dict.team_id through the gate.
---
 litellm/proxy/common_request_processing.py    |  8 ++++---
 .../proxy/test_common_request_processing.py   | 21 +++++++++++++++++--
 2 files changed, 24 insertions(+), 5 deletions(-)

diff --git a/litellm/proxy/common_request_processing.py b/litellm/proxy/common_request_processing.py
index 07dea5268d8..1ca14ea3684 100644
--- a/litellm/proxy/common_request_processing.py
+++ b/litellm/proxy/common_request_processing.py
@@ -258,7 +258,7 @@ def _litellm_model_supports_stream_options(litellm_model: str) -> bool:
     return supported_params is not None and "stream_options" in supported_params
 
 
-def _deployment_litellm_model(deployment: Mapping[str, object]) -> Optional[str]:
+def _deployment_litellm_model(deployment: Mapping[str, object]) -> str | None:
     litellm_params = deployment.get("litellm_params")
     if isinstance(litellm_params, Mapping):
         litellm_model = litellm_params.get("model")
@@ -269,11 +269,12 @@ def _deployment_litellm_model(deployment: Mapping[str, object]) -> Optional[str]
 
 def _model_deployments_support_stream_options(
     model: object,
-    llm_router: Optional[Router],
+    llm_router: Router | None,
+    team_id: str | None,
 ) -> bool:
     if not isinstance(model, str):
         return False
-    deployments = llm_router.get_model_list(model_name=model) if llm_router is not None else None
+    deployments = llm_router.get_model_list(model_name=model, team_id=team_id) if llm_router is not None else None
     deployment_models = tuple(
         litellm_model
         for deployment in deployments or ()
@@ -1309,6 +1310,7 @@ class ProxyBaseLLMRequestProcessing:
                 supports_stream_options=lambda: _model_deployments_support_stream_options(
                     model=self.data.get("model"),
                     llm_router=llm_router,
+                    team_id=user_api_key_dict.team_id,
                 ),
             )
         )
diff --git a/tests/test_litellm/proxy/test_common_request_processing.py b/tests/test_litellm/proxy/test_common_request_processing.py
index 28fb97006d8..3bb84e095a0 100644
--- a/tests/test_litellm/proxy/test_common_request_processing.py
+++ b/tests/test_litellm/proxy/test_common_request_processing.py
@@ -5265,12 +5265,12 @@ class TestApplyStreamUsageTracking:
 
 
 class TestModelDeploymentsSupportStreamOptions:
-    def _support(self, model, llm_router=None) -> bool:
+    def _support(self, model, llm_router=None, team_id=None) -> bool:
         from litellm.proxy.common_request_processing import (
             _model_deployments_support_stream_options,
         )
 
-        return _model_deployments_support_stream_options(model=model, llm_router=llm_router)
+        return _model_deployments_support_stream_options(model=model, llm_router=llm_router, team_id=team_id)
 
     def test_openai_compatible_deployment_supports_stream_options(self):
         router = litellm.Router(
@@ -5335,5 +5335,22 @@ class TestModelDeploymentsSupportStreamOptions:
     def test_unmapped_model_name_is_not_injected(self):
         assert self._support("some-unmapped-public-alias", None) is False
 
+    def test_team_alias_model_resolves_with_team_id(self):
+        router = litellm.Router(
+            model_list=[
+                {
+                    "model_name": "model_name_team-1_8b6a0b3f",
+                    "litellm_params": {"model": "azure/gpt-5.4-nano", "api_key": "fake"},
+                    "model_info": {
+                        "team_id": "team-1",
+                        "team_public_model_name": "team-gpt",
+                    },
+                }
+            ]
+        )
+
+        assert self._support("team-gpt", router, team_id="team-1") is True
+        assert self._support("team-gpt", router, team_id=None) is False
+
     def test_non_string_model_is_not_injected(self):
         assert self._support(None, None) is False