diff --git a/litellm/proxy/auth/model_checks.py b/litellm/proxy/auth/model_checks.py
index 6f91b1693d2..50f319d0121 100644
--- a/litellm/proxy/auth/model_checks.py
+++ b/litellm/proxy/auth/model_checks.py
@@ -64,30 +64,26 @@ def _get_models_from_access_groups(
     return all_models
 
 
-def _is_access_group_only_string(
+def _is_unresolvable_model_identifier(
     model: str,
     proxy_model_list: List[str],
     model_access_groups: Dict[str, List[str]],
 ) -> bool:
     """
-    Identify entries in a key/team `models` list that are access-group strings
-    rather than real model identifiers.
-
-    Issue #25550: when a virtual key was created with `models=["foo"]` against
-    an access group "foo" that has since been removed (or that was never
-    defined on any model), "foo" remains in the key but is no longer in
-    `model_access_groups`. The current expansion logic only strips entries
-    that ARE in the access-groups dict, so the stale string passes through
-    /v1/models as if it were a model name.
-
-    A string is treated as access-group-only when none of the markers of a
-    real model identifier are present:
-      - it is in proxy_model_list (a configured proxy model group)
+    A string from a key/team `models` list resolves to a real model id when
+    any of these are true:
+      - it is in proxy_model_list (a configured proxy model name, including
+        custom enterprise aliases not present in litellm.model_list_set)
       - it is in model_access_groups (an active access group; expanded above)
       - it is in litellm.model_list_set (a known base model id)
       - it contains "/" (provider-qualified route, e.g. `openai/gpt-4o`)
       - it contains "*" (wildcard route, expanded by `_get_wildcard_models`)
       - it starts with "ft:" (OpenAI fine-tune id)
+
+    Anything else is unresolvable and would otherwise leak into /v1/models as
+    if it were a real model. The most common case (issue #25550) is a stale
+    access-group label whose group was removed; identifiers for proxy models
+    that have since been removed from config are filtered for the same reason.
     """
     if model in proxy_model_list:
         return False
@@ -249,14 +245,16 @@ def get_complete_model_list(
             valid_models = get_valid_models()
             append_unique(valid_models)
 
-    # Drop stale access-group strings carried in key/team `models` (issue #25550).
-    # Only filter the key/team paths — the proxy-admin path above is sourced
-    # from authoritative state (proxy_model_list, model_access_groups keys).
+    # Drop unresolvable identifiers carried in key/team `models` — most
+    # commonly stale access-group labels (issue #25550), but also names of
+    # proxy models that have since been removed from config. Only filter the
+    # key/team paths; the proxy-admin path above is sourced from authoritative
+    # state (proxy_model_list, model_access_groups keys).
     if key_models or team_models:
         unique_models = [
             m
             for m in unique_models
-            if not _is_access_group_only_string(
+            if not _is_unresolvable_model_identifier(
                 model=m,
                 proxy_model_list=proxy_model_list,
                 model_access_groups=model_access_groups,
diff --git a/tests/test_litellm/proxy/auth/test_model_checks.py b/tests/test_litellm/proxy/auth/test_model_checks.py
index f2f30a88738..efbe628100f 100644
--- a/tests/test_litellm/proxy/auth/test_model_checks.py
+++ b/tests/test_litellm/proxy/auth/test_model_checks.py
@@ -338,6 +338,28 @@ def test_get_complete_model_list_keeps_known_base_model():
     assert result == ["gpt-4o-mini"]
 
 
+def test_get_complete_model_list_keeps_custom_proxy_alias():
+    """
+    A custom enterprise proxy model name (e.g. 'internal-assistant') that is
+    listed in proxy_model_list but is NOT a known LiteLLM base model id must
+    survive the filter. This locks in the proxy_model_list-membership branch
+    of _is_unresolvable_model_identifier — the most common preservation path
+    in real deployments.
+    """
+    from litellm.proxy.auth.model_checks import get_complete_model_list
+
+    result = get_complete_model_list(
+        key_models=["internal-assistant"],
+        team_models=[],
+        proxy_model_list=["internal-assistant"],
+        user_model=None,
+        infer_model_from_keys=False,
+        model_access_groups={},
+    )
+
+    assert result == ["internal-assistant"]
+
+
 def test_get_complete_model_list_keeps_finetune_id():
     """OpenAI fine-tune ids (`ft:...`) must survive the filter."""
     from litellm.proxy.auth.model_checks import get_complete_model_list