diff --git a/litellm/proxy/proxy_server.py b/litellm/proxy/proxy_server.py
index 63109916ab1..e4db02758a1 100644
--- a/litellm/proxy/proxy_server.py
+++ b/litellm/proxy/proxy_server.py
@@ -8139,17 +8139,26 @@ def _add_team_models_to_all_models(
                     if can_add_model:
                         team_models.setdefault(model_id, set()).add(team_object.team_id)
         else:
+            model_access_groups = (
+                llm_router.get_model_access_groups(team_id=team_object.team_id)
+                if llm_router
+                else {}
+            )
             for model_name in team_object.models:
-                _models = llm_router.get_model_list(
-                    model_name=model_name, team_id=team_object.team_id
-                )
-                if _models is not None:
-                    for model in _models:
-                        model_id = model.get("model_info", {}).get("id", None)
-                        if model_id is not None:
-                            team_models.setdefault(model_id, set()).add(
-                                team_object.team_id
-                            )
+                names_to_resolve = [model_name]
+                if model_name in model_access_groups:
+                    names_to_resolve = model_access_groups[model_name]
+                for name in names_to_resolve:
+                    _models = llm_router.get_model_list(
+                        model_name=name, team_id=team_object.team_id
+                    )
+                    if _models is not None:
+                        for model in _models:
+                            model_id = model.get("model_info", {}).get("id", None)
+                            if model_id is not None:
+                                team_models.setdefault(model_id, set()).add(
+                                    team_object.team_id
+                                )
     return team_models
 
 
@@ -8702,18 +8711,25 @@ async def _filter_models_by_team_id(
                 if can_add_model:
                     team_accessible_model_ids.add(model_id)
     else:
-        # Team has access to specific models
+        # Team has access to specific models (or model access group names like "on-prem")
+        model_access_groups = (
+            llm_router.get_model_access_groups(team_id=team_id) if llm_router else {}
+        )
         for model_name in team_object.models:
-            _models = (
-                llm_router.get_model_list(model_name=model_name, team_id=team_id)
-                if llm_router
-                else []
-            )
-            if _models is not None:
-                for model in _models:
-                    model_id = model.get("model_info", {}).get("id", None)
-                    if model_id is not None:
-                        team_accessible_model_ids.add(model_id)
+            names_to_resolve = [model_name]
+            if model_name in model_access_groups:
+                names_to_resolve = model_access_groups[model_name]
+            for name in names_to_resolve:
+                _models = (
+                    llm_router.get_model_list(model_name=name, team_id=team_id)
+                    if llm_router
+                    else []
+                )
+                if _models is not None:
+                    for model in _models:
+                        model_id = model.get("model_info", {}).get("id", None)
+                        if model_id is not None:
+                            team_accessible_model_ids.add(model_id)
 
     # Also search database for models accessible to this team
     # This complements the config search done above
diff --git a/tests/test_litellm/proxy/test_proxy_server.py b/tests/test_litellm/proxy/test_proxy_server.py
index d65df0087ad..c3e8d4429cd 100644
--- a/tests/test_litellm/proxy/test_proxy_server.py
+++ b/tests/test_litellm/proxy/test_proxy_server.py
@@ -896,6 +896,37 @@ def test_add_team_models_to_all_models():
     assert result == {"gpt-4-model-2": {"team1"}}
 
 
+def test_add_team_models_to_all_models_resolves_access_groups():
+    """
+    Team with models = ["on-prem"] (access group name) should get model ids
+    for models in that access group.
+    """
+    from litellm.proxy._types import LiteLLM_TeamTable
+    from litellm.proxy.proxy_server import _add_team_models_to_all_models
+
+    team = MagicMock(spec=LiteLLM_TeamTable)
+    team.team_id = "team-1"
+    team.models = ["on-prem"]
+
+    llm_router = MagicMock()
+    llm_router.get_model_access_groups.return_value = {
+        "on-prem": ["gemma3-4b", "qwen3-embedding-8b"],
+    }
+    llm_router.get_model_list.side_effect = lambda model_name, team_id: [
+        {"model_info": {"id": f"id-{model_name}", "team_id": None}},
+    ]
+
+    result = _add_team_models_to_all_models(
+        team_db_objects_typed=[team],
+        llm_router=llm_router,
+    )
+
+    assert result == {
+        "id-gemma3-4b": {"team-1"},
+        "id-qwen3-embedding-8b": {"team-1"},
+    }
+
+
 @pytest.mark.asyncio
 async def test_delete_deployment_type_mismatch():
     """