scriptoriumadmin/litellm
1
0
Fork 0
mirror of https://github.com/BerriAI/litellm.git synced 2026-11-03 07:23:33 +00:00
Code Issues Projects Releases Packages Wiki Activity Actions 166

fix(openrouter): keep /v1/models cheap when no wildcard catalog is cached

Browse source
This commit is contained in:
mateo-berri 2026-10-30 05:45:34 -07:00
parent 9176e4df18
commit 2b91f3be98
1 changed files with 7 additions and 1 deletions
Show stats Download patch file Download diff file Expand all files Collapse all files

8
litellm/router.py
View file

@ -1737,6 +1737,7 @@ class Router(DeprecatedRouterMembers):
max_size_in_memory=max(len(model_list or ()), 1),
default_ttl=2 * MODEL_INFO_REFRESH_SECONDS,
)
self._has_wildcard_catalog: bool = False
self._routing_group_rows: tuple[DeploymentTypedDict, ...] | None = None
self._init_routing_groups(None)
self._provider_unresolved_deployments: tuple[Callable[[], Deployment | None], ...] = ()
@ -12273,6 +12274,7 @@ class Router(DeprecatedRouterMembers):
self._discovered_model_info_cache.max_size_in_memory = max(len(self._core_model_list), 1)
self._discovered_model_info_cache.delete_cache(model_id)
self._discovered_model_info_cache.set_cache(model_id, discovered)
self._has_wildcard_catalog = self._has_wildcard_catalog or bool(discovered.limits_by_model)
self._invalidate_model_group_info_cache()
async def _adiscover_openai_compatible_model_info(
@ -12423,7 +12425,11 @@ class Router(DeprecatedRouterMembers):
)
def _wildcard_catalog_model_infos(self, model_name: str) -> Iterator[Mapping[str, object]]:
for deployment, routed_model in self._core_pattern_router.get_routed_deployments(model_name):
if not self._has_wildcard_catalog:
return
for deployment, routed_model in self._core_pattern_router.get_routed_deployments(
model_name, LlmProviders.OPENROUTER.value
):
configured = _configured_model_info(deployment)
model_id = configured.get("id")
discovered = self._core_get_discovered_model_info(