Merge pull request #5620 from BerriAI/litellm_use_helper_to_get_httpx_clients

[Feat-Perf] Use common helper to get async httpx clients for all providers

docs/my-website/docs/proxy/prod.md

@@ -107,9 +107,9 @@ export LITELLM_SALT_KEY="sk-1234"
 |--------------|-------|
 | Avg latency | `50ms` |
 | Median latency | `51ms` |
-| `/chat/completions` Requests/second | `35` |
+| `/chat/completions` Requests/second | `100` |
-| `/chat/completions` Requests/minute | `2100` |
+| `/chat/completions` Requests/minute | `6000` |
-| `/chat/completions` Requests/hour | `126K` |
+| `/chat/completions` Requests/hour | `360K` |
 
 
 ### Verifying Debugging logs are off

enterprise/enterprise_hooks/aporia_ai.py

@@ -26,7 +26,10 @@ from typing import List
 from datetime import datetime
 import aiohttp, asyncio
 from litellm._logging import verbose_proxy_logger
-from litellm.llms.custom_httpx.http_handler import AsyncHTTPHandler
+from litellm.llms.custom_httpx.http_handler import (
+    get_async_httpx_client,
+    httpxSpecialProvider,
+)
 import httpx
 import json
 from litellm.types.guardrails import GuardrailEventHooks
@@ -40,8 +43,8 @@ class AporiaGuardrail(CustomGuardrail):
     def __init__(
         self, api_key: Optional[str] = None, api_base: Optional[str] = None, **kwargs
     ):
-        self.async_handler = AsyncHTTPHandler(
+        self.async_handler = get_async_httpx_client(
-            timeout=httpx.Timeout(timeout=600.0, connect=5.0)
+            llm_provider=httpxSpecialProvider.GuardrailCallback
         )
         self.aporia_api_key = api_key or os.environ["APORIO_API_KEY"]
         self.aporia_api_base = api_base or os.environ["APORIO_API_BASE"]

litellm/__init__.py

@@ -24,6 +24,7 @@ from litellm.proxy._types import (
 )
 import httpx
 import dotenv
+from enum import Enum
 
 litellm_mode = os.getenv("LITELLM_MODE", "DEV")  # "PRODUCTION", "DEV"
 if litellm_mode == "DEV":
@@ -678,62 +679,66 @@ model_list = (
     + gemini_models
 )
 
-provider_list: List = [
+
-    "openai",
+class LlmProviders(str, Enum):
-    "custom_openai",
+    OPENAI = "openai"
-    "text-completion-openai",
+    CUSTOM_OPENAI = "custom_openai"
-    "cohere",
+    TEXT_COMPLETION_OPENAI = "text-completion-openai"
-    "cohere_chat",
+    COHERE = "cohere"
-    "clarifai",
+    COHERE_CHAT = "cohere_chat"
-    "anthropic",
+    CLARIFAI = "clarifai"
-    "replicate",
+    ANTHROPIC = "anthropic"
-    "huggingface",
+    REPLICATE = "replicate"
-    "together_ai",
+    HUGGINGFACE = "huggingface"
-    "openrouter",
+    TOGETHER_AI = "together_ai"
-    "vertex_ai",
+    OPENROUTER = "openrouter"
-    "vertex_ai_beta",
+    VERTEX_AI = "vertex_ai"
-    "palm",
+    VERTEX_AI_BETA = "vertex_ai_beta"
-    "gemini",
+    PALM = "palm"
-    "ai21",
+    GEMINI = "gemini"
-    "baseten",
+    AI21 = "ai21"
-    "azure",
+    BASETEN = "baseten"
-    "azure_text",
+    AZURE = "azure"
-    "azure_ai",
+    AZURE_TEXT = "azure_text"
-    "sagemaker",
+    AZURE_AI = "azure_ai"
-    "sagemaker_chat",
+    SAGEMAKER = "sagemaker"
-    "bedrock",
+    SAGEMAKER_CHAT = "sagemaker_chat"
-    "vllm",
+    BEDROCK = "bedrock"
-    "nlp_cloud",
+    VLLM = "vllm"
-    "petals",
+    NLP_CLOUD = "nlp_cloud"
-    "oobabooga",
+    PETALS = "petals"
-    "ollama",
+    OOBABOOGA = "oobabooga"
-    "ollama_chat",
+    OLLAMA = "ollama"
-    "deepinfra",
+    OLLAMA_CHAT = "ollama_chat"
-    "perplexity",
+    DEEPINFRA = "deepinfra"
-    "anyscale",
+    PERPLEXITY = "perplexity"
-    "mistral",
+    ANYSCALE = "anyscale"
-    "groq",
+    MISTRAL = "mistral"
-    "nvidia_nim",
+    GROQ = "groq"
-    "cerebras",
+    NVIDIA_NIM = "nvidia_nim"
-    "ai21_chat",
+    CEREBRAS = "cerebras"
-    "volcengine",
+    AI21_CHAT = "ai21_chat"
-    "codestral",
+    VOLCENGINE = "volcengine"
-    "text-completion-codestral",
+    CODESTRAL = "codestral"
-    "deepseek",
+    TEXT_COMPLETION_CODESTRAL = "text-completion-codestral"
-    "maritalk",
+    DEEPSEEK = "deepseek"
-    "voyage",
+    MARITALK = "maritalk"
-    "cloudflare",
+    VOYAGE = "voyage"
-    "xinference",
+    CLOUDFLARE = "cloudflare"
-    "fireworks_ai",
+    XINFERENCE = "xinference"
-    "friendliai",
+    FIREWORKS_AI = "fireworks_ai"
-    "watsonx",
+    FRIENDLIAI = "friendliai"
-    "triton",
+    WATSONX = "watsonx"
-    "predibase",
+    TRITON = "triton"
-    "databricks",
+    PREDIBASE = "predibase"
-    "empower",
+    DATABRICKS = "databricks"
-    "github",
+    EMPOWER = "empower"
-    "custom",  # custom apis
+    GITHUB = "github"
-]
+    CUSTOM = "custom"
+
+
+provider_list: List[Union[LlmProviders, str]] = list(LlmProviders)
+
 
 models_by_provider: dict = {
     "openai": open_ai_chat_completion_models + open_ai_text_completion_models,

litellm/caching.py

@@ -1242,8 +1242,9 @@ class QdrantSemanticCache(BaseCache):
         import os
 
         from litellm.llms.custom_httpx.http_handler import (
-            _get_async_httpx_client,
             _get_httpx_client,
+            get_async_httpx_client,
+            httpxSpecialProvider,
         )
 
         if collection_name is None:
@@ -1290,7 +1291,9 @@ class QdrantSemanticCache(BaseCache):
         self.headers = headers
 
         self.sync_client = _get_httpx_client()
-        self.async_client = _get_async_httpx_client()
+        self.async_client = get_async_httpx_client(
+            llm_provider=httpxSpecialProvider.Caching
+        )
 
         if quantization_config is None:
             print_verbose(

litellm/integrations/braintrust_logging.py

@@ -17,10 +17,17 @@ from pydantic import BaseModel
 import litellm
 from litellm import verbose_logger
 from litellm.integrations.custom_logger import CustomLogger
-from litellm.llms.custom_httpx.http_handler import AsyncHTTPHandler, HTTPHandler
+from litellm.llms.custom_httpx.http_handler import (
+    AsyncHTTPHandler,
+    HTTPHandler,
+    get_async_httpx_client,
+    httpxSpecialProvider,
+)
 from litellm.utils import get_formatted_prompt
 
-global_braintrust_http_handler = AsyncHTTPHandler()
+global_braintrust_http_handler = get_async_httpx_client(
+    llm_provider=httpxSpecialProvider.LoggingCallback
+)
 global_braintrust_sync_http_handler = HTTPHandler()
 API_BASE = "https://api.braintrustdata.com/v1"
 

litellm/integrations/galileo.py

@@ -8,7 +8,10 @@ from pydantic import BaseModel, Field
 import litellm
 from litellm._logging import verbose_logger
 from litellm.integrations.custom_logger import CustomLogger
-from litellm.llms.custom_httpx.http_handler import AsyncHTTPHandler
+from litellm.llms.custom_httpx.http_handler import (
+    get_async_httpx_client,
+    httpxSpecialProvider,
+)
 
 
 # from here: https://docs.rungalileo.io/galileo/gen-ai-studio-products/galileo-observe/how-to/logging-data-via-restful-apis#structuring-your-records
@@ -39,8 +42,8 @@ class GalileoObserve(CustomLogger):
         self.base_url = os.getenv("GALILEO_BASE_URL", None)
         self.project_id = os.getenv("GALILEO_PROJECT_ID", None)
         self.headers = None
-        self.async_httpx_handler = AsyncHTTPHandler(
+        self.async_httpx_handler = get_async_httpx_client(
-            timeout=httpx.Timeout(timeout=600.0, connect=5.0)
+            llm_provider=httpxSpecialProvider.LoggingCallback
         )
         pass
 

litellm/integrations/gcs_bucket_base.py

@@ -13,15 +13,18 @@ from litellm.integrations.custom_logger import CustomLogger
 from litellm.litellm_core_utils.logging_utils import (
     convert_litellm_response_object_to_dict,
 )
-from litellm.llms.custom_httpx.http_handler import AsyncHTTPHandler
+from litellm.llms.custom_httpx.http_handler import (
+    get_async_httpx_client,
+    httpxSpecialProvider,
+)
 
 
 class GCSBucketBase(CustomLogger):
     def __init__(self, bucket_name: Optional[str] = None) -> None:
         from litellm.proxy.proxy_server import premium_user
 
-        self.async_httpx_client = AsyncHTTPHandler(
+        self.async_httpx_client = get_async_httpx_client(
-            timeout=httpx.Timeout(timeout=600.0, connect=5.0)
+            llm_provider=httpxSpecialProvider.LoggingCallback
         )
         self.path_service_account_json = os.getenv("GCS_PATH_SERVICE_ACCOUNT", None)
         self.BUCKET_NAME = bucket_name or os.getenv("GCS_BUCKET_NAME", None)

litellm/integrations/lago.py

@@ -13,7 +13,11 @@ import httpx
 import litellm
 from litellm import verbose_logger
 from litellm.integrations.custom_logger import CustomLogger
-from litellm.llms.custom_httpx.http_handler import AsyncHTTPHandler, HTTPHandler
+from litellm.llms.custom_httpx.http_handler import (
+    HTTPHandler,
+    get_async_httpx_client,
+    httpxSpecialProvider,
+)
 
 
 def get_utc_datetime():
@@ -30,7 +34,9 @@ class LagoLogger(CustomLogger):
     def __init__(self) -> None:
         super().__init__()
         self.validate_environment()
-        self.async_http_handler = AsyncHTTPHandler()
+        self.async_http_handler = get_async_httpx_client(
+            llm_provider=httpxSpecialProvider.LoggingCallback
+        )
         self.sync_http_handler = HTTPHandler()
 
     def validate_environment(self):

litellm/integrations/langsmith.py

@@ -16,7 +16,11 @@ from pydantic import BaseModel  # type: ignore
 import litellm
 from litellm._logging import verbose_logger
 from litellm.integrations.custom_logger import CustomLogger
-from litellm.llms.custom_httpx.http_handler import AsyncHTTPHandler
+from litellm.llms.custom_httpx.http_handler import (
+    AsyncHTTPHandler,
+    get_async_httpx_client,
+    httpxSpecialProvider,
+)
 
 
 class LangsmithInputs(BaseModel):
@@ -61,8 +65,8 @@ class LangsmithLogger(CustomLogger):
         self.langsmith_base_url = os.getenv(
             "LANGSMITH_BASE_URL", "https://api.smith.langchain.com"
         )
-        self.async_httpx_client = AsyncHTTPHandler(
+        self.async_httpx_client = get_async_httpx_client(
-            timeout=httpx.Timeout(timeout=600.0, connect=5.0)
+            llm_provider=httpxSpecialProvider.LoggingCallback
         )
 
     def _prepare_log_data(self, kwargs, response_obj, start_time, end_time):

litellm/integrations/openmeter.py

@@ -12,7 +12,12 @@ import httpx
 import litellm
 from litellm import verbose_logger
 from litellm.integrations.custom_logger import CustomLogger
-from litellm.llms.custom_httpx.http_handler import AsyncHTTPHandler, HTTPHandler
+from litellm.llms.custom_httpx.http_handler import (
+    AsyncHTTPHandler,
+    HTTPHandler,
+    get_async_httpx_client,
+    httpxSpecialProvider,
+)
 
 
 def get_utc_datetime():
@@ -29,7 +34,9 @@ class OpenMeterLogger(CustomLogger):
     def __init__(self) -> None:
         super().__init__()
         self.validate_environment()
-        self.async_http_handler = AsyncHTTPHandler()
+        self.async_http_handler = get_async_httpx_client(
+            llm_provider=httpxSpecialProvider.LoggingCallback
+        )
         self.sync_http_handler = HTTPHandler()
 
     def validate_environment(self):

litellm/integrations/prometheus_helpers/prometheus_api.py

@@ -8,11 +8,16 @@ import time
 
 import litellm
 from litellm._logging import verbose_logger
-from litellm.llms.custom_httpx.http_handler import AsyncHTTPHandler
+from litellm.llms.custom_httpx.http_handler import (
+    get_async_httpx_client,
+    httpxSpecialProvider,
+)
 
 PROMETHEUS_URL = litellm.get_secret("PROMETHEUS_URL")
 PROMETHEUS_SELECTED_INSTANCE = litellm.get_secret("PROMETHEUS_SELECTED_INSTANCE")
-async_http_handler = AsyncHTTPHandler()
+async_http_handler = get_async_httpx_client(
+    llm_provider=httpxSpecialProvider.LoggingCallback
+)
 
 
 async def get_metric_from_prometheus(

litellm/integrations/slack_alerting.py

@@ -25,7 +25,11 @@ from litellm._logging import verbose_logger, verbose_proxy_logger
 from litellm.caching import DualCache
 from litellm.integrations.custom_logger import CustomLogger
 from litellm.litellm_core_utils.litellm_logging import Logging
-from litellm.llms.custom_httpx.http_handler import AsyncHTTPHandler
+from litellm.llms.custom_httpx.http_handler import (
+    AsyncHTTPHandler,
+    get_async_httpx_client,
+    httpxSpecialProvider,
+)
 from litellm.proxy._types import (
     AlertType,
     CallInfo,
@@ -187,7 +191,9 @@ class SlackAlerting(CustomLogger):
         self.alerting = alerting
         self.alert_types = alert_types
         self.internal_usage_cache = internal_usage_cache or DualCache()
-        self.async_http_handler = AsyncHTTPHandler()
+        self.async_http_handler = get_async_httpx_client(
+            llm_provider=httpxSpecialProvider.LoggingCallback
+        )
         self.alert_to_webhook_url = alert_to_webhook_url
         self.is_running = False
         self.alerting_args = SlackAlertingArgs(**alerting_args)

litellm/llms/anthropic/chat.py

@@ -25,8 +25,8 @@ from litellm.litellm_core_utils.core_helpers import map_finish_reason
 from litellm.llms.custom_httpx.http_handler import (
     AsyncHTTPHandler,
     HTTPHandler,
-    _get_async_httpx_client,
     _get_httpx_client,
+    get_async_httpx_client,
 )
 from litellm.types.llms.anthropic import (
     AnthopicMessagesAssistantMessageParam,
@@ -918,7 +918,9 @@ class AnthropicChatCompletion(BaseLLM):
         headers={},
         client=None,
     ) -> Union[ModelResponse, CustomStreamWrapper]:
-        async_handler = _get_async_httpx_client()
+        async_handler = get_async_httpx_client(
+            llm_provider=litellm.LlmProviders.ANTHROPIC
+        )
 
         try:
             response = await async_handler.post(

litellm/llms/bedrock/chat.py

@@ -35,8 +35,8 @@ from litellm.litellm_core_utils.litellm_logging import Logging
 from litellm.llms.custom_httpx.http_handler import (
     AsyncHTTPHandler,
     HTTPHandler,
-    _get_async_httpx_client,
     _get_httpx_client,
+    get_async_httpx_client,
 )
 from litellm.types.llms.bedrock import *
 from litellm.types.llms.openai import (
@@ -209,7 +209,9 @@ async def make_call(
 ):
     try:
         if client is None:
-            client = _get_async_httpx_client()  # Create a new client if none provided
+            client = get_async_httpx_client(
+                llm_provider=litellm.LlmProviders.BEDROCK
+            )  # Create a new client if none provided
 
         response = await client.post(
             api_base,
@@ -1041,7 +1043,7 @@ class BedrockLLM(BaseAWSLLM):
                 if isinstance(timeout, float) or isinstance(timeout, int):
                     timeout = httpx.Timeout(timeout)
                 _params["timeout"] = timeout
-            client = _get_async_httpx_client(_params)  # type: ignore
+            client = get_async_httpx_client(params=_params, llm_provider=litellm.LlmProviders.BEDROCK)  # type: ignore
         else:
             client = client  # type: ignore
 
@@ -1498,7 +1500,9 @@ class BedrockConverseLLM(BaseAWSLLM):
                 if isinstance(timeout, float) or isinstance(timeout, int):
                     timeout = httpx.Timeout(timeout)
                 _params["timeout"] = timeout
-            client = _get_async_httpx_client(_params)  # type: ignore
+            client = get_async_httpx_client(
+                params=_params, llm_provider=litellm.LlmProviders.BEDROCK
+            )
         else:
             client = client  # type: ignore
 

litellm/llms/bedrock/embed/embedding.py

@@ -15,8 +15,8 @@ from litellm.llms.cohere.embed import embedding as cohere_embedding
 from litellm.llms.custom_httpx.http_handler import (
     AsyncHTTPHandler,
     HTTPHandler,
-    _get_async_httpx_client,
     _get_httpx_client,
+    get_async_httpx_client,
 )
 from litellm.secret_managers.main import get_secret
 from litellm.types.llms.bedrock import AmazonEmbeddingRequest, CohereEmbeddingRequest
@@ -130,7 +130,9 @@ class BedrockEmbedding(BaseAWSLLM):
                 if isinstance(timeout, float) or isinstance(timeout, int):
                     timeout = httpx.Timeout(timeout)
                 _params["timeout"] = timeout
-            client = _get_async_httpx_client(_params)  # type: ignore
+            client = get_async_httpx_client(
+                params=_params, llm_provider=litellm.LlmProviders.BEDROCK
+            )
         else:
             client = client
 

litellm/llms/cohere/rerank.py

@@ -9,10 +9,11 @@ from typing import Any, Dict, List, Optional, Union
 import httpx
 from pydantic import BaseModel
 
+import litellm
 from litellm.llms.base import BaseLLM
 from litellm.llms.custom_httpx.http_handler import (
-    _get_async_httpx_client,
     _get_httpx_client,
+    get_async_httpx_client,
 )
 from litellm.rerank_api.types import RerankRequest, RerankResponse
 
@@ -65,7 +66,7 @@ class CohereRerank(BaseLLM):
         api_key: str,
         api_base: str,
     ) -> RerankResponse:
-        client = _get_async_httpx_client()
+        client = get_async_httpx_client(llm_provider=litellm.LlmProviders.COHERE)
 
         response = await client.post(
             api_base,

litellm/llms/custom_httpx/http_handler.py

@@ -1,12 +1,19 @@
 import asyncio
 import os
 import traceback
-from typing import Any, Mapping, Optional, Union
+from typing import TYPE_CHECKING, Any, Mapping, Optional, Union
 
 import httpx
 
 import litellm
 
+from .types import httpxSpecialProvider
+
+if TYPE_CHECKING:
+    from litellm import LlmProviders
+else:
+    LlmProviders = Any
+
 try:
     from litellm._version import version
 except:
@@ -378,7 +385,10 @@ class HTTPHandler:
             pass
 
 
-def _get_async_httpx_client(params: Optional[dict] = None) -> AsyncHTTPHandler:
+def get_async_httpx_client(
+    llm_provider: Union[LlmProviders, httpxSpecialProvider],
+    params: Optional[dict] = None,
+) -> AsyncHTTPHandler:
     """
     Retrieves the async HTTP client from the cache
     If not present, creates a new client
@@ -393,7 +403,7 @@ def _get_async_httpx_client(params: Optional[dict] = None) -> AsyncHTTPHandler:
             except Exception:
                 pass
 
-    _cache_key_name = "async_httpx_client" + _params_key_name
+    _cache_key_name = "async_httpx_client" + _params_key_name + llm_provider
     if _cache_key_name in litellm.in_memory_llm_clients_cache:
         return litellm.in_memory_llm_clients_cache[_cache_key_name]
 

litellm/llms/custom_httpx/types.py

@@ -0,0 +1,10 @@
+from enum import Enum
+
+import litellm
+
+
+class httpxSpecialProvider(str, Enum):
+    LoggingCallback = "logging_callback"
+    GuardrailCallback = "guardrail_callback"
+    Caching = "caching"
+    Oauth2Check = "oauth2_check"

litellm/llms/sagemaker/sagemaker.py

@@ -19,8 +19,8 @@ from litellm.litellm_core_utils.asyncify import asyncify
 from litellm.llms.custom_httpx.http_handler import (
     AsyncHTTPHandler,
     HTTPHandler,
-    _get_async_httpx_client,
     _get_httpx_client,
+    get_async_httpx_client,
 )
 from litellm.types.llms.openai import (
     ChatCompletionToolCallChunk,
@@ -565,8 +565,8 @@ class SagemakerLLM(BaseAWSLLM):
     ):
         try:
             if client is None:
-                client = (
+                client = get_async_httpx_client(
-                    _get_async_httpx_client()
+                    llm_provider=litellm.LlmProviders.SAGEMAKER
                 )  # Create a new client if none provided
             response = await client.post(
                 api_base,
@@ -673,7 +673,9 @@ class SagemakerLLM(BaseAWSLLM):
         model_id: Optional[str],
     ):
         timeout = 300.0
-        async_handler = _get_async_httpx_client()
+        async_handler = get_async_httpx_client(
+            llm_provider=litellm.LlmProviders.SAGEMAKER
+        )
 
         async_transform_prompt = asyncify(self._transform_prompt)
 

litellm/llms/togetherai/rerank.py

@@ -9,10 +9,11 @@ from typing import Any, Dict, List, Optional, Union
 import httpx
 from pydantic import BaseModel
 
+import litellm
 from litellm.llms.base import BaseLLM
 from litellm.llms.custom_httpx.http_handler import (
-    _get_async_httpx_client,
     _get_httpx_client,
+    get_async_httpx_client,
 )
 from litellm.rerank_api.types import RerankRequest, RerankResponse
 
@@ -77,7 +78,9 @@ class TogetherAIRerank(BaseLLM):
         request_data_dict: Dict[str, Any],
         api_key: str,
     ) -> RerankResponse:
-        client = _get_async_httpx_client()  # Use async client
+        client = get_async_httpx_client(
+            llm_provider=litellm.LlmProviders.TOGETHER_AI
+        )  # Use async client
 
         response = await client.post(
             "https://api.together.xyz/v1/rerank",

litellm/llms/vertex_ai_and_google_ai_studio/gemini/vertex_and_google_ai_studio_gemini.py

@@ -22,7 +22,7 @@ from litellm.litellm_core_utils.core_helpers import map_finish_reason
 from litellm.llms.custom_httpx.http_handler import (
     AsyncHTTPHandler,
     HTTPHandler,
-    _get_async_httpx_client,
+    get_async_httpx_client,
 )
 from litellm.llms.prompt_templates.factory import (
     convert_url_to_base64,
@@ -1294,7 +1294,9 @@ class VertexLLM(BaseLLM):
         if timeout:
             _async_client_params["timeout"] = timeout
         if client is None or not isinstance(client, AsyncHTTPHandler):
-            client = _get_async_httpx_client(params=_async_client_params)
+            client = get_async_httpx_client(
+                params=_async_client_params, llm_provider=litellm.LlmProviders.VERTEX_AI
+            )
         else:
             client = client  # type: ignore
         ## LOGGING

litellm/llms/vertex_ai_and_google_ai_studio/text_to_speech/text_to_speech_handler.py

@@ -4,13 +4,14 @@ from typing import Any, Coroutine, Literal, Optional, TypedDict, Union
 
 import httpx
 
+import litellm
 from litellm._logging import verbose_logger
 from litellm.llms.base import BaseLLM
 from litellm.llms.custom_httpx.http_handler import (
     AsyncHTTPHandler,
     HTTPHandler,
-    _get_async_httpx_client,
     _get_httpx_client,
+    get_async_httpx_client,
 )
 from litellm.llms.OpenAI.openai import HttpxBinaryResponseContent
 from litellm.llms.vertex_ai_and_google_ai_studio.gemini.vertex_and_google_ai_studio_gemini import (
@@ -178,7 +179,9 @@ class VertexTextToSpeechAPI(VertexLLM):
     ) -> HttpxBinaryResponseContent:
         import base64
 
-        async_handler = _get_async_httpx_client()
+        async_handler = get_async_httpx_client(
+            llm_provider=litellm.LlmProviders.VERTEX_AI
+        )
 
         response = await async_handler.post(
             url=url,

litellm/proxy/auth/oauth2_check.py

@@ -20,7 +20,10 @@ async def check_oauth2_token(token: str) -> UserAPIKeyAuth:
     import httpx
 
     from litellm._logging import verbose_proxy_logger
-    from litellm.llms.custom_httpx.http_handler import _get_async_httpx_client
+    from litellm.llms.custom_httpx.http_handler import (
+        get_async_httpx_client,
+        httpxSpecialProvider,
+    )
     from litellm.proxy._types import CommonProxyErrors
     from litellm.proxy.proxy_server import premium_user
 
@@ -40,7 +43,7 @@ async def check_oauth2_token(token: str) -> UserAPIKeyAuth:
     if not token_info_endpoint:
         raise ValueError("OAUTH_TOKEN_INFO_ENDPOINT environment variable is not set")
 
-    client = _get_async_httpx_client()
+    client = get_async_httpx_client(llm_provider=httpxSpecialProvider.Oauth2Check)
     headers = {"Authorization": f"Bearer {token}", "Content-Type": "application/json"}
 
     try:

litellm/proxy/guardrails/guardrail_hooks/aporia_ai.py

@@ -30,7 +30,11 @@ from litellm.integrations.custom_guardrail import CustomGuardrail
 from litellm.litellm_core_utils.logging_utils import (
     convert_litellm_response_object_to_str,
 )
-from litellm.llms.custom_httpx.http_handler import AsyncHTTPHandler
+from litellm.llms.custom_httpx.http_handler import (
+    AsyncHTTPHandler,
+    get_async_httpx_client,
+    httpxSpecialProvider,
+)
 from litellm.proxy._types import UserAPIKeyAuth
 from litellm.proxy.guardrails.guardrail_helpers import should_proceed_based_on_metadata
 from litellm.types.guardrails import GuardrailEventHooks
@@ -44,8 +48,8 @@ class AporiaGuardrail(CustomGuardrail):
     def __init__(
         self, api_key: Optional[str] = None, api_base: Optional[str] = None, **kwargs
     ):
-        self.async_handler = AsyncHTTPHandler(
+        self.async_handler = get_async_httpx_client(
-            timeout=httpx.Timeout(timeout=600.0, connect=5.0)
+            llm_provider=httpxSpecialProvider.GuardrailCallback
         )
         self.aporia_api_key = api_key or os.environ["APORIO_API_KEY"]
         self.aporia_api_base = api_base or os.environ["APORIO_API_BASE"]

litellm/proxy/guardrails/guardrail_hooks/bedrock_guardrails.py

@@ -33,7 +33,8 @@ from litellm.litellm_core_utils.logging_utils import (
 from litellm.llms.base_aws_llm import BaseAWSLLM
 from litellm.llms.custom_httpx.http_handler import (
     AsyncHTTPHandler,
-    _get_async_httpx_client,
+    get_async_httpx_client,
+    httpxSpecialProvider,
 )
 from litellm.proxy._types import UserAPIKeyAuth
 from litellm.proxy.guardrails.guardrail_helpers import should_proceed_based_on_metadata
@@ -55,7 +56,9 @@ class BedrockGuardrail(CustomGuardrail, BaseAWSLLM):
         guardrailVersion: Optional[str] = None,
         **kwargs,
     ):
-        self.async_handler = _get_async_httpx_client()
+        self.async_handler = get_async_httpx_client(
+            llm_provider=httpxSpecialProvider.GuardrailCallback
+        )
         self.guardrailIdentifier = guardrailIdentifier
         self.guardrailVersion = guardrailVersion
 

litellm/proxy/guardrails/guardrail_hooks/lakera_ai.py

@@ -21,7 +21,11 @@ from fastapi import HTTPException
 import litellm
 from litellm._logging import verbose_proxy_logger
 from litellm.integrations.custom_guardrail import CustomGuardrail
-from litellm.llms.custom_httpx.http_handler import AsyncHTTPHandler
+from litellm.llms.custom_httpx.http_handler import (
+    AsyncHTTPHandler,
+    get_async_httpx_client,
+    httpxSpecialProvider,
+)
 from litellm.proxy._types import UserAPIKeyAuth
 from litellm.proxy.guardrails.guardrail_helpers import should_proceed_based_on_metadata
 from litellm.secret_managers.main import get_secret
@@ -50,8 +54,8 @@ class lakeraAI_Moderation(CustomGuardrail):
         api_key: Optional[str] = None,
         **kwargs,
     ):
-        self.async_handler = AsyncHTTPHandler(
+        self.async_handler = get_async_httpx_client(
-            timeout=httpx.Timeout(timeout=600.0, connect=5.0)
+            llm_provider=httpxSpecialProvider.GuardrailCallback
         )
         self.lakera_api_key = api_key or os.environ["LAKERA_API_KEY"]
         self.moderation_check = moderation_check

litellm/proxy/guardrails/guardrail_hooks/presidio.py

@@ -22,7 +22,6 @@ import litellm  # noqa: E401
 from litellm._logging import verbose_proxy_logger
 from litellm.caching import DualCache
 from litellm.integrations.custom_guardrail import CustomGuardrail
-from litellm.llms.custom_httpx.http_handler import _get_async_httpx_client
 from litellm.proxy._types import UserAPIKeyAuth
 from litellm.utils import (
     EmbeddingResponse,