fix(logging): classify async anthropic_messages and generate_content as async (#33589)

litellm/google_genai/main.py

@@ -17,6 +17,7 @@ from litellm.llms.base_llm.google_genai.transformation import (
 )
 from litellm.llms.custom_httpx.llm_http_handler import BaseLLMHTTPHandler
 from litellm.types.router import GenericLiteLLMParams
+from litellm.types.utils import CallTypes
 from litellm.utils import ProviderConfigManager, client
 
 if TYPE_CHECKING:
@@ -39,6 +40,11 @@ base_llm_http_handler = BaseLLMHTTPHandler()
 #################################################
 
 
+def _mark_async_entrypoint(logging_obj: LiteLLMLoggingObj | None, marker: str, is_async: bool) -> None:
+    if logging_obj is not None:
+        logging_obj.model_call_details.setdefault("litellm_params", {})[marker] = is_async
+
+
 class GenerateContentSetupResult(BaseModel):
     """Internal Type - Result of setting up a generate content call"""
 
@@ -315,6 +321,8 @@ def generate_content(
     try:
         _is_async = kwargs.pop("agenerate_content", False)
 
+        _mark_async_entrypoint(kwargs.get("litellm_logging_obj"), CallTypes.agenerate_content.value, _is_async)
+
         # Handle generationConfig parameter from kwargs for backward compatibility
         if "generationConfig" in kwargs and config is None:
             config = kwargs.pop("generationConfig")
@@ -403,6 +411,8 @@ async def agenerate_content_stream(
     try:
         kwargs["agenerate_content_stream"] = True
 
+        _mark_async_entrypoint(kwargs.get("litellm_logging_obj"), CallTypes.agenerate_content_stream.value, True)
+
         # Handle generationConfig parameter from kwargs for backward compatibility
         if "generationConfig" in kwargs and config is None:
             config = kwargs.pop("generationConfig")
@@ -497,6 +507,8 @@ def generate_content_stream(
         # Remove any async-related flags since this is the sync function
         _is_async = kwargs.pop("agenerate_content_stream", False)
 
+        _mark_async_entrypoint(kwargs.get("litellm_logging_obj"), CallTypes.agenerate_content_stream.value, _is_async)
+
         # Handle generationConfig parameter from kwargs for backward compatibility
         if "generationConfig" in kwargs and config is None:
             config = kwargs.pop("generationConfig")

litellm/litellm_core_utils/litellm_logging.py

@@ -1531,6 +1531,9 @@ class Logging(LiteLLMLoggingBaseClass):
             and litellm_params.get(CallTypes.aimage_generation.value, False) is not True
             and litellm_params.get(CallTypes.atranscription.value, False) is not True
             and litellm_params.get(CallTypes.allm_passthrough_route.value, False) is not True
+            and litellm_params.get(CallTypes.aanthropic_messages.value, False) is not True
+            and litellm_params.get(CallTypes.agenerate_content.value, False) is not True
+            and litellm_params.get(CallTypes.agenerate_content_stream.value, False) is not True
         )
 
     def _is_assembled_stream_success(self, result=None) -> bool:

litellm/llms/anthropic/experimental_pass_through/messages/handler.py

@@ -36,6 +36,7 @@ from litellm.types.llms.anthropic_messages.anthropic_response import (
     AnthropicMessagesResponse,
 )
 from litellm.types.router import GenericLiteLLMParams
+from litellm.types.utils import CallTypes
 from litellm.utils import ProviderConfigManager, client
 
 from ..utils import is_reasoning_auto_summary_enabled
@@ -463,6 +464,9 @@ def anthropic_messages_handler(
             "model": original_model,
             "custom_llm_provider": custom_llm_provider,
         }
+        litellm_logging_obj.model_call_details.setdefault("litellm_params", {})[CallTypes.aanthropic_messages.value] = (
+            is_async
+        )
 
         # Check if stream was converted for WebSearch interception
         # This is set in the async wrapper above when stream=True is converted to stream=False

litellm/types/utils.py

@@ -328,6 +328,7 @@ class CallTypes(str, Enum):
     cancel_batch = "cancel_batch"
     pass_through = "pass_through_endpoint"
     anthropic_messages = "anthropic_messages"
+    aanthropic_messages = "aanthropic_messages"
     get_assistants = "get_assistants"
     aget_assistants = "aget_assistants"
     create_assistants = "create_assistants"
@@ -496,6 +497,7 @@ CallTypesLiteral = Literal[
     "pass_through_endpoint",
     "allm_passthrough_route",
     "anthropic_messages",
+    "aanthropic_messages",
     "aretrieve_batch",
     "retrieve_batch",
     "generate_content",

tests/test_litellm/litellm_core_utils/test_litellm_logging.py

@@ -653,6 +653,80 @@ async def test_logging_result_for_bridge_calls(logging_obj):
         assert mock_should_run_logging.call_count == 1
 
 
+@pytest.mark.asyncio
+async def test_anthropic_messages_marks_litellm_params_async():
+    """LIT-4447: the async ``anthropic_messages`` entrypoint must plant
+    ``aanthropic_messages`` in ``litellm_params`` so ``_is_sync_litellm_request``
+    classifies the request async and the sync CustomLogger hook does not fire in
+    addition to the async one, mirroring how ``acompletion`` / ``aresponses`` set
+    their own async markers."""
+    import asyncio
+
+    import litellm
+    from litellm.integrations.custom_logger import CustomLogger
+
+    captured = {}
+    logged = asyncio.Event()
+
+    class CaptureLogger(CustomLogger):
+        async def async_log_success_event(self, kwargs, response_obj, start_time, end_time):
+            captured["litellm_params"] = kwargs.get("litellm_params", {})
+            logged.set()
+
+    logger = CaptureLogger()
+    logger.log_success_event = MagicMock()
+    original_callbacks = getattr(litellm, "callbacks", [])
+    try:
+        litellm.callbacks = [logger]
+        await litellm.anthropic_messages(
+            max_tokens=100,
+            messages=[{"role": "user", "content": "Hey"}],
+            model="anthropic/claude-sonnet-4-5",
+            mock_response="Hello, world!",
+        )
+        await asyncio.wait_for(logged.wait(), timeout=10)
+
+        assert captured["litellm_params"].get("aanthropic_messages") is True
+        assert LitellmLogging._is_sync_litellm_request(captured["litellm_params"]) is False
+        logger.log_success_event.assert_not_called()
+    finally:
+        litellm.callbacks = original_callbacks
+
+
+@pytest.mark.asyncio
+async def test_agenerate_content_marks_litellm_params_async():
+    """LIT-4475: the async ``agenerate_content`` entrypoint must plant
+    ``agenerate_content`` in ``litellm_params`` so ``_is_sync_litellm_request``
+    classifies the nested delegated call async, preventing the sync CustomLogger
+    hook from firing alongside the async one."""
+    import time
+
+    import litellm
+
+    logging_obj = LitellmLogging(
+        model="gemini/gemini-2.0-flash",
+        messages=[{"role": "user", "content": "hi"}],
+        stream=False,
+        call_type="agenerate_content",
+        start_time=time.time(),
+        litellm_call_id="agenerate-content-marker-check",
+        function_id="fn",
+    )
+    try:
+        await litellm.agenerate_content(
+            model="gemini/gemini-2.0-flash",
+            contents=[{"role": "user", "parts": [{"text": "hi"}]}],
+            mock_response="hello",
+            litellm_logging_obj=logging_obj,
+        )
+    except Exception:
+        pass
+
+    litellm_params = logging_obj.model_call_details.get("litellm_params", {})
+    assert litellm_params.get("agenerate_content") is True
+    assert LitellmLogging._is_sync_litellm_request(litellm_params) is False
+
+
 @pytest.mark.asyncio
 async def test_logging_non_streaming_request():
     import asyncio
@@ -712,7 +786,15 @@ async def test_logging_non_streaming_request():
 
 
 @pytest.mark.parametrize(
-    "async_flag", ["acompletion", "aresponses", "allm_passthrough_route"]
+    "async_flag",
+    [
+        "acompletion",
+        "aresponses",
+        "allm_passthrough_route",
+        "aanthropic_messages",
+        "agenerate_content",
+        "agenerate_content_stream",
+    ],
 )
 def test_success_handler_skips_sync_callbacks_for_async_requests(
     logging_obj, async_flag
@@ -805,6 +887,17 @@ def test_is_sync_litellm_request():
         LitellmLogging._is_sync_litellm_request({"allm_passthrough_route": True})
         is False
     )
+    assert (
+        LitellmLogging._is_sync_litellm_request({"aanthropic_messages": True}) is False
+    )
+    assert LitellmLogging._is_sync_litellm_request({"agenerate_content": True}) is False
+    assert (
+        LitellmLogging._is_sync_litellm_request({"agenerate_content_stream": True})
+        is False
+    )
+    assert (
+        LitellmLogging._is_sync_litellm_request({"aanthropic_messages": False}) is True
+    )
 
 
 def test_get_litellm_params_propagates_allm_passthrough_route():

tests/test_litellm/llms/azure/test_azure_common_utils.py

@@ -426,6 +426,7 @@ def test_select_azure_base_url_called(setup_mocks):
             "arerank",
             "arealtime",
             "anthropic_messages",
+            "aanthropic_messages",
             "add_message",
             "arun_thread_stream",
             "aresponses",

ui/litellm-dashboard/src/lib/http/schema.d.ts

@@ -21690,7 +21690,7 @@ export interface components {
          * CallTypes
          * @enum {string}
          */
-        CallTypes: "embedding" | "aembedding" | "completion" | "acompletion" | "atext_completion" | "text_completion" | "image_generation" | "aimage_generation" | "image_edit" | "aimage_edit" | "moderation" | "amoderation" | "atranscription" | "transcription" | "aspeech" | "speech" | "rerank" | "arerank" | "search" | "asearch" | "_arealtime" | "_aresponses_websocket" | "create_batch" | "acreate_batch" | "aretrieve_batch" | "retrieve_batch" | "acancel_batch" | "cancel_batch" | "pass_through_endpoint" | "anthropic_messages" | "get_assistants" | "aget_assistants" | "create_assistants" | "acreate_assistants" | "delete_assistant" | "adelete_assistant" | "acreate_thread" | "create_thread" | "aget_thread" | "get_thread" | "a_add_message" | "add_message" | "aget_messages" | "get_messages" | "arun_thread" | "run_thread" | "arun_thread_stream" | "run_thread_stream" | "afile_retrieve" | "file_retrieve" | "afile_delete" | "file_delete" | "afile_list" | "file_list" | "acreate_file" | "create_file" | "afile_content" | "file_content" | "create_fine_tuning_job" | "acreate_fine_tuning_job" | "create_video" | "acreate_video" | "avideo_retrieve" | "video_retrieve" | "avideo_content" | "video_content" | "video_remix" | "avideo_remix" | "video_list" | "avideo_list" | "video_retrieve_job" | "avideo_retrieve_job" | "video_delete" | "avideo_delete" | "video_create_character" | "avideo_create_character" | "video_get_character" | "avideo_get_character" | "video_edit" | "avideo_edit" | "video_extension" | "avideo_extension" | "vector_store_file_create" | "avector_store_file_create" | "vector_store_file_list" | "avector_store_file_list" | "vector_store_file_retrieve" | "avector_store_file_retrieve" | "vector_store_file_content" | "avector_store_file_content" | "vector_store_file_update" | "avector_store_file_update" | "vector_store_file_delete" | "avector_store_file_delete" | "vector_store_create" | "avector_store_create" | "vector_store_search" | "avector_store_search" | "create_container" | "acreate_container" | "list_containers" | "alist_containers" | "retrieve_container" | "aretrieve_container" | "delete_container" | "adelete_container" | "list_container_files" | "alist_container_files" | "upload_container_file" | "aupload_container_file" | "create_sandbox" | "acreate_sandbox" | "delete_sandbox" | "adelete_sandbox" | "run_code" | "arun_code" | "code_interpreter_tool" | "acode_interpreter_tool" | "acancel_fine_tuning_job" | "cancel_fine_tuning_job" | "alist_fine_tuning_jobs" | "list_fine_tuning_jobs" | "aretrieve_fine_tuning_job" | "retrieve_fine_tuning_job" | "responses" | "aresponses" | "alist_input_items" | "llm_passthrough_route" | "allm_passthrough_route" | "generate_content" | "agenerate_content" | "generate_content_stream" | "agenerate_content_stream" | "ocr" | "aocr" | "call_mcp_tool" | "list_mcp_tools" | "asend_message" | "send_message" | "acreate_skill";
+        CallTypes: "embedding" | "aembedding" | "completion" | "acompletion" | "atext_completion" | "text_completion" | "image_generation" | "aimage_generation" | "image_edit" | "aimage_edit" | "moderation" | "amoderation" | "atranscription" | "transcription" | "aspeech" | "speech" | "rerank" | "arerank" | "search" | "asearch" | "_arealtime" | "_aresponses_websocket" | "create_batch" | "acreate_batch" | "aretrieve_batch" | "retrieve_batch" | "acancel_batch" | "cancel_batch" | "pass_through_endpoint" | "anthropic_messages" | "aanthropic_messages" | "get_assistants" | "aget_assistants" | "create_assistants" | "acreate_assistants" | "delete_assistant" | "adelete_assistant" | "acreate_thread" | "create_thread" | "aget_thread" | "get_thread" | "a_add_message" | "add_message" | "aget_messages" | "get_messages" | "arun_thread" | "run_thread" | "arun_thread_stream" | "run_thread_stream" | "afile_retrieve" | "file_retrieve" | "afile_delete" | "file_delete" | "afile_list" | "file_list" | "acreate_file" | "create_file" | "afile_content" | "file_content" | "create_fine_tuning_job" | "acreate_fine_tuning_job" | "create_video" | "acreate_video" | "avideo_retrieve" | "video_retrieve" | "avideo_content" | "video_content" | "video_remix" | "avideo_remix" | "video_list" | "avideo_list" | "video_retrieve_job" | "avideo_retrieve_job" | "video_delete" | "avideo_delete" | "video_create_character" | "avideo_create_character" | "video_get_character" | "avideo_get_character" | "video_edit" | "avideo_edit" | "video_extension" | "avideo_extension" | "vector_store_file_create" | "avector_store_file_create" | "vector_store_file_list" | "avector_store_file_list" | "vector_store_file_retrieve" | "avector_store_file_retrieve" | "vector_store_file_content" | "avector_store_file_content" | "vector_store_file_update" | "avector_store_file_update" | "vector_store_file_delete" | "avector_store_file_delete" | "vector_store_create" | "avector_store_create" | "vector_store_search" | "avector_store_search" | "create_container" | "acreate_container" | "list_containers" | "alist_containers" | "retrieve_container" | "aretrieve_container" | "delete_container" | "adelete_container" | "list_container_files" | "alist_container_files" | "upload_container_file" | "aupload_container_file" | "create_sandbox" | "acreate_sandbox" | "delete_sandbox" | "adelete_sandbox" | "run_code" | "arun_code" | "code_interpreter_tool" | "acode_interpreter_tool" | "acancel_fine_tuning_job" | "cancel_fine_tuning_job" | "alist_fine_tuning_jobs" | "list_fine_tuning_jobs" | "aretrieve_fine_tuning_job" | "retrieve_fine_tuning_job" | "responses" | "aresponses" | "alist_input_items" | "llm_passthrough_route" | "allm_passthrough_route" | "generate_content" | "agenerate_content" | "generate_content_stream" | "agenerate_content_stream" | "ocr" | "aocr" | "call_mcp_tool" | "list_mcp_tools" | "asend_message" | "send_message" | "acreate_skill";
         /** CallbackDelete */
         CallbackDelete: {
             /** Callback Name */