diff --git a/litellm/proxy/auth/auth_utils.py b/litellm/proxy/auth/auth_utils.py index 5ff85ad4a0e..b6c4a43149d 100644 --- a/litellm/proxy/auth/auth_utils.py +++ b/litellm/proxy/auth/auth_utils.py @@ -980,12 +980,6 @@ def get_key_own_model_rate_limit( user_api_key_dict: UserAPIKeyAuth, rate_limit_key: Literal["model_rpm_limit", "model_tpm_limit"], ) -> dict[str, int] | None: - """ - Per-model limit the key sets on itself: key metadata first, then model_max_budget. - - Unlike get_key_model_rpm_limit / get_key_model_tpm_limit this never falls back to the - team, so callers can tell a key override apart from an inherited team limit. - """ if user_api_key_dict.metadata: result: Final = user_api_key_dict.metadata.get(rate_limit_key) if result: diff --git a/litellm/proxy/hooks/parallel_request_limiter_v3.py b/litellm/proxy/hooks/parallel_request_limiter_v3.py index 87336830976..f35fb1e0042 100644 --- a/litellm/proxy/hooks/parallel_request_limiter_v3.py +++ b/litellm/proxy/hooks/parallel_request_limiter_v3.py @@ -2899,7 +2899,6 @@ class _PROXY_MaxParallelRequestsHandler_v3(CustomLogger): requested_model: str, rate_limit_key: Literal["model_rpm_limit", "model_tpm_limit"], ) -> int | None: - """Team per-model limit this key inherits: None when the key sets its own limit for the model.""" team_limits: Final = get_model_rate_limit_from_metadata(user_api_key_dict, "team_metadata", rate_limit_key) team_limit: Final = team_limits.get(requested_model) if team_limits else None if team_limit is None: @@ -2915,7 +2914,6 @@ class _PROXY_MaxParallelRequestsHandler_v3(CustomLogger): requested_model: str | None, descriptors: list[RateLimitDescriptor], ) -> None: - """Add the team's per-model descriptor for the metrics the key does not override itself.""" if requested_model is None: return team_rpm_limit: Final = self._inherited_team_model_limit(user_api_key_dict, requested_model, "model_rpm_limit")