fix(vector_stores): run the full model grant check on caller-supplied model hints

The vector-store file routes accept a model hint through the ?model= query
param and the x-litellm-model header. That hint was authorized with a hand
rolled check that covered only the key allowlist and the team allowlist, so
a key restricted by its project's model grant, a team-member restriction, or a
key config still routed through the hinted deployment. Greptile flagged the
gap as a P1 on the replacement PR.

The hint now goes through the same authorize_model_for_key path the batches
and files routes use, which runs can_key_call_resolved_model with every rule
the proxy enforces elsewhere. Keys those extra rules deny now get a 403 on
these routes. The two remaining behavioral differences are edge cases the
old check tolerated: a key whose team_models is set without a team_id no
longer runs the team allowlist, and a key with a config set skips the key
allowlist, both matching the rest of the proxy.

The regression test caches a project whose grant excludes the hinted model
and asserts the request is refused before any deployment lookup. The two
patch() calls on litellm.proxy.proxy_server carry a test-quality-ok reason
because can_key_call_resolved_model reads prisma_client and
user_api_key_cache through a lazy module import with no injection seam.

litellm/proxy/vector_store_files_endpoints/endpoints.py

@@ -5,7 +5,6 @@ from fastapi.responses import ORJSONResponse
 
 import litellm
 from litellm.proxy._types import UserAPIKeyAuth
-from litellm.proxy.auth.auth_checks import _can_object_call_model, can_key_call_model
 from litellm.proxy.auth.user_api_key_auth import user_api_key_auth
 from litellm.proxy.common_request_processing import ProxyBaseLLMRequestProcessing
 from litellm.proxy.common_utils.openai_endpoint_utils import (
@@ -14,6 +13,7 @@ from litellm.proxy.common_utils.openai_endpoint_utils import (
     get_custom_llm_provider_from_request_query,
 )
 from litellm.proxy.openai_files_endpoints.common_utils import (
+    authorize_model_for_key,
     get_credentials_for_model,
     handle_model_based_routing,
     prepare_data_with_credentials,
@@ -212,26 +212,7 @@ async def _authorize_model_routing_hint(
 ) -> None:
     if user_api_key_dict is None:
         return
-
-    key_models: Final = getattr(user_api_key_dict, "models", None)
-    if not (isinstance(key_models, list) and "all-team-models" in key_models):
-        await can_key_call_model(
-            model=model,
-            llm_model_list=None,
-            valid_token=user_api_key_dict,
-            llm_router=llm_router,
-        )
-
-    team_models: Final = getattr(user_api_key_dict, "team_models", None)
-    if isinstance(team_models, list) and len(team_models) > 0:
-        _can_object_call_model(
-            model=model,
-            llm_router=llm_router,
-            models=team_models,
-            team_model_aliases=user_api_key_dict.team_model_aliases,
-            team_id=user_api_key_dict.team_id,
-            object_type="team",
-        )
+    await authorize_model_for_key(model_id=model, llm_router=llm_router, user_api_key_dict=user_api_key_dict)
 
 
 async def _update_request_data_with_model_routing_hint(

tests/test_litellm/proxy/vector_store_endpoints/test_vector_store_endpoints.py

@@ -609,6 +609,44 @@ async def test_vector_store_file_list_authorizes_model_query_param_before_creden
     llm_router.get_deployment_credentials_with_provider.assert_not_called()
 
 
+@pytest.mark.asyncio
+async def test_vector_store_file_list_model_query_param_enforces_project_model_grant():
+    from litellm.proxy._types import LiteLLM_ProjectTableCachedObj, LiteLLM_TeamTableCachedObj
+    from litellm.proxy.auth.auth_checks import ProxyException
+    from litellm.proxy.common_utils.user_api_key_cache import UserApiKeyCache, project_cache_key
+
+    request = MagicMock(spec=Request)
+    request.query_params = {"model": "team-openai"}
+    request.headers = {}
+
+    llm_router = MagicMock()
+    llm_router.model_group_alias = {}
+    cache = UserApiKeyCache()
+    await cache.async_set_cache(
+        key="team_id:team-123",
+        value=LiteLLM_TeamTableCachedObj(team_id="team-123", models=["team-openai"]),
+    )
+    await cache.async_set_cache(
+        key=project_cache_key("proj-1"),
+        value=LiteLLM_ProjectTableCachedObj(project_id="proj-1", models=["other-deployment"]),
+    )
+    user_api_key_dict = UserAPIKeyAuth(team_id="team-123", team_models=["team-openai"], project_id="proj-1")
+
+    with (
+        patch("litellm.proxy.proxy_server.prisma_client", MagicMock()),  # test-quality-ok: proxy_server global, no seam
+        patch("litellm.proxy.proxy_server.user_api_key_cache", cache),  # test-quality-ok: proxy_server global, no seam
+    ):
+        with pytest.raises(ProxyException):
+            await _update_request_data_with_model_routing_hint(
+                data={"vector_store_id": "vs_123"},
+                request=request,
+                llm_router=llm_router,
+                user_api_key_dict=user_api_key_dict,
+            )
+
+    llm_router.get_deployment_credentials_with_provider.assert_not_called()
+
+
 @pytest.mark.asyncio
 async def test_update_request_data_with_litellm_managed_vector_store_registry():
     """