diff --git a/litellm/responses/streaming_iterator.py b/litellm/responses/streaming_iterator.py
index 58f55ac73fd..33adb663021 100644
--- a/litellm/responses/streaming_iterator.py
+++ b/litellm/responses/streaming_iterator.py
@@ -1899,6 +1899,12 @@ async def _admit_frame_turn(turn_admission: Callable[[], Awaitable[None]] | None
     await turn_admission()
 
 
+def _turn_refusal_frame(refusal: RateLimitError) -> str:
+    return json.dumps(
+        {"type": "error", "status": 429, "error": {"type": "rate_limit_exceeded", "message": str(refusal)}}
+    )
+
+
 RESPONSES_WS_LOGGED_EVENT_TYPES: Final = [
     "response.created",
     "response.completed",
@@ -2547,17 +2553,7 @@ class ResponsesWebSocketStreaming:
             await _admit_frame_turn(turn_admission, message)
         except RateLimitError as e:
             try:
-                await self.websocket.send_text(
-                    json.dumps(
-                        {
-                            "type": "error",
-                            "error": {
-                                "type": "rate_limit_exceeded",
-                                "message": str(e),
-                            },
-                        }
-                    )
-                )
+                await self.websocket.send_text(_turn_refusal_frame(e))
             except Exception:  # noqa: BLE001, S110  # client may already be gone
                 pass
             return False
@@ -2702,10 +2698,11 @@ class ManagedResponsesWebSocketHandler:
             return None
 
     async def _send_error(self, message: str, error_type: str = "server_error") -> None:
+        await self._send_frame(json.dumps({"type": "error", "error": {"type": error_type, "message": message}}))
+
+    async def _send_frame(self, frame: str) -> None:
         try:
-            await self.websocket.send_text(
-                json.dumps({"type": "error", "error": {"type": error_type, "message": message}})
-            )
+            await self.websocket.send_text(frame)
         except Exception:
             pass
 
@@ -3085,7 +3082,7 @@ class ManagedResponsesWebSocketHandler:
             )
             await _admit_frame_turn(turn_admission, raw_message)
         except RateLimitError as e:
-            await self._send_error(str(e), error_type="rate_limit_exceeded")
+            await self._send_frame(_turn_refusal_frame(e))
             return
 
         call_kwargs: Final = self._build_base_call_kwargs(msg_obj)
diff --git a/tests/unit/responses/test_responses_websocket_all_providers.py b/tests/unit/responses/test_responses_websocket_all_providers.py
index 92c815aa4ab..1d80d87a75d 100644
--- a/tests/unit/responses/test_responses_websocket_all_providers.py
+++ b/tests/unit/responses/test_responses_websocket_all_providers.py
@@ -1388,6 +1388,7 @@ class TestWebSocketTurnAdmission:
         ]
         assert turn_admission.await_count == 2
         refusal: Final = json.loads(websocket.send_text.await_args.args[0])
+        assert refusal["status"] == 429
         assert refusal["error"]["type"] == "rate_limit_exceeded"
         assert "crossed budget" in refusal["error"]["message"]
 
@@ -1442,7 +1443,7 @@ class TestWebSocketTurnAdmission:
             for call in websocket.send_text.await_args_list
             if json.loads(call.args[0]).get("type") == "error"
         ]
-        assert [refusal["error"]["type"] for refusal in refusals] == ["rate_limit_exceeded"]
+        assert [(refusal["status"], refusal["error"]["type"]) for refusal in refusals] == [(429, "rate_limit_exceeded")]
 
 
 def _deployment_defaults():