perf: avoid unconditional router lookups in success handler

Replace bare _get_deployment_default_tpm/rpm_limit calls in the
async_log_success_event condition with get_key_model_tpm/rpm_limit
(model_name=model_group). The higher-level getters short-circuit on
key/team metadata hits before ever reaching the router, so requests
that don't use deployment defaults incur no extra router lookup. Remove
the now-unused bare helper imports.

Also fix invalid `int = None` type hints in test helper signatures
to `Optional[int] = None`.

Co-Authored-By: Claude (claude-sonnet-4-6) <noreply@anthropic.com>

litellm/proxy/hooks/parallel_request_limiter.py

@@ -14,8 +14,6 @@ from litellm.integrations.custom_logger import CustomLogger
 from litellm.litellm_core_utils.core_helpers import _get_parent_otel_span_from_kwargs
 from litellm.proxy._types import CommonProxyErrors, CurrentItemRateLimit, UserAPIKeyAuth
 from litellm.proxy.auth.auth_utils import (
-    _get_deployment_default_rpm_limit,
-    _get_deployment_default_tpm_limit,
     get_key_model_rpm_limit,
     get_key_model_tpm_limit,
 )
@@ -548,8 +546,14 @@ class _PROXY_MaxParallelRequestsHandler(CustomLogger):
                     "model_rpm_limit" in user_api_key_metadata
                     or "model_tpm_limit" in user_api_key_metadata
                     or user_api_key_model_max_budget is not None
-                    or _get_deployment_default_tpm_limit(model_group) is not None
-                    or _get_deployment_default_rpm_limit(model_group) is not None
+                    or get_key_model_tpm_limit(
+                        user_api_key_dict, model_name=model_group
+                    )
+                    is not None
+                    or get_key_model_rpm_limit(
+                        user_api_key_dict, model_name=model_group
+                    )
+                    is not None
                 )
             ):
                 request_count_api_key = (

tests/test_litellm/proxy/auth/test_auth_utils.py

@@ -2,6 +2,7 @@
 Unit tests for auth_utils functions related to rate limiting and customer ID extraction.
 """
 
+from typing import Optional
 from unittest.mock import MagicMock, patch
 
 from litellm.proxy._types import UserAPIKeyAuth
@@ -317,7 +318,7 @@ def test_get_end_user_id_falls_back_to_deprecated_user_header_name():
     assert result == "user-legacy"
 
 
-def _make_deployment_dict(model_name: str, tpm: int = None, rpm: int = None) -> dict:
+def _make_deployment_dict(model_name: str, tpm: Optional[int] = None, rpm: Optional[int] = None) -> dict:
     """Helper to build a minimal deployment dict as returned by router.get_model_list."""
     litellm_params: dict = {"model": model_name}
     if tpm is not None:

tests/test_litellm/proxy/test_model_info_default_limits.py

@@ -3,6 +3,7 @@ Tests verifying that default_api_key_tpm_limit and default_api_key_rpm_limit set
 litellm_params are returned by the /model/info endpoint.
 """
 
+from typing import Optional
 from unittest.mock import MagicMock, patch
 
 import pytest
@@ -13,8 +14,8 @@ from litellm.types.router import Deployment, LiteLLM_Params, ModelInfo
 
 def _make_deployment(
     model_name: str,
-    default_tpm: int = None,
-    default_rpm: int = None,
+    default_tpm: Optional[int] = None,
+    default_rpm: Optional[int] = None,
 ) -> Deployment:
     params: dict = {"model": f"openai/{model_name}"}
     if default_tpm is not None: