From b86900da0979d01e7ffc69e3fa9cbb1eb6ba263c Mon Sep 17 00:00:00 2001
From: mateo-berri <277851410+mateo-berri@users.noreply.github.com>
Date: Tue, 20 Oct 2026 21:54:43 -0700
Subject: [PATCH] fix(proxy): free the max_parallel_requests slot when a
 replayed /v1/messages MCP stream completes

A streamed /v1/messages request that auto-executes MCP tools runs its model
rounds without streaming and replays the final answer as a stream. The first
round claims the request's only success log, and its limiter callback keeps the
slot because the tool call reused it afterwards. The stream path released the
slot only on a client disconnect, so after message_stop the key stayed at its
max_parallel_requests limit and its next request answered 429

Stream cleanup now releases the stashed slot when the stream completed and its
logging object already claimed the success log, which only happens when the
stream replays a response that was logged before it started. A real stream
still releases through its own success callback
---
 litellm/litellm_core_utils/litellm_logging.py |  4 ++++
 litellm/proxy/common_request_processing.py    | 10 +++++++++-
 2 files changed, 13 insertions(+), 1 deletion(-)

diff --git a/litellm/litellm_core_utils/litellm_logging.py b/litellm/litellm_core_utils/litellm_logging.py
index a59672b5cb2..b5714989cc8 100644
--- a/litellm/litellm_core_utils/litellm_logging.py
+++ b/litellm/litellm_core_utils/litellm_logging.py
@@ -2583,6 +2583,10 @@ class Logging(LiteLLMLoggingBaseClass):
         self._async_success_scheduled = True
         return True
 
+    @property
+    def async_success_claimed(self) -> bool:
+        return self._async_success_scheduled
+
     def mark_logging_complete(
         self,
         event_type: Literal["async_success", "sync_success", "async_failure", "sync_failure"],
diff --git a/litellm/proxy/common_request_processing.py b/litellm/proxy/common_request_processing.py
index 6fda5780f63..6557e5088a3 100644
--- a/litellm/proxy/common_request_processing.py
+++ b/litellm/proxy/common_request_processing.py
@@ -4353,6 +4353,15 @@ class ProxyBaseLLMRequestProcessing:
                     and user_api_key_dict is not None
                 ):
                     await proxy_logging_obj.arelease_max_parallel_requests_on_disconnect(user_api_key_dict)
+            logging_obj: Final = request_data.get("litellm_logging_obj")
+            if (
+                stream_completed
+                and isinstance(logging_obj, LiteLLMLoggingObj)
+                and logging_obj.async_success_claimed
+                and proxy_logging_obj is not None
+                and user_api_key_dict is not None
+            ):
+                await proxy_logging_obj.arelease_max_parallel_requests_on_disconnect(user_api_key_dict)
 
             if hasattr(response, "aclose"):
                 try:
@@ -4362,7 +4371,6 @@ class ProxyBaseLLMRequestProcessing:
                         "async_streaming_data_generator: error closing response stream: %s",
                         e,
                     )
-            logging_obj: Final = request_data.get("litellm_logging_obj")
             if (
                 not stream_completed
                 and isinstance(logging_obj, LiteLLMLoggingObj)