fix anthropic_messages implementation

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

@@ -6,7 +6,7 @@
 """
 
 import json
-from typing import Any, AsyncIterator, Dict, Optional, Union, cast
+from typing import AsyncIterator, Dict, List, Optional, Union, cast
 
 import httpx
 
@@ -64,9 +64,20 @@ class AnthropicMessagesHandler:
 
 @client
 async def anthropic_messages(
-    api_key: str,
+    max_tokens: int,
+    messages: List[Dict[str, Union[str, List[Dict[str, str]]]]],
     model: str,
-    stream: bool = False,
+    metadata: Optional[Dict] = None,
+    stop_sequences: Optional[List[str]] = None,
+    stream: Optional[bool] = False,
+    system: Optional[str] = None,
+    temperature: Optional[float] = 1.0,
+    thinking: Optional[Dict] = None,
+    tool_choice: Optional[Dict] = None,
+    tools: Optional[List[Dict]] = None,
+    top_k: Optional[int] = None,
+    top_p: Optional[float] = None,
+    api_key: Optional[str] = None,
     api_base: Optional[str] = None,
     client: Optional[AsyncHTTPHandler] = None,
     custom_llm_provider: Optional[str] = None,
@@ -133,7 +144,7 @@ async def anthropic_messages(
     litellm_logging_obj.model_call_details.update(kwargs)
 
     # Prepare request body
-    request_body = kwargs.copy()
+    request_body = locals().copy()
     request_body = {
         k: v
         for k, v in request_body.items()
@@ -165,7 +176,7 @@ async def anthropic_messages(
         url=request_url,
         headers=headers,
         data=json.dumps(request_body),
-        stream=stream,
+        stream=stream or False,
     )
     response.raise_for_status()
 

litellm/messages/__init__.py

@@ -4,13 +4,105 @@ Interface for Anthropic's messages API
 Use this to call LLMs in Anthropic /messages Request/Response format
 """
 
+from typing import Dict, List, Optional, Union
+
 from litellm.llms.anthropic.experimental_pass_through.handler import (
     anthropic_messages as _async_anthropic_messages,
 )
 
 
-async def acreate(*args, **kwargs):
+async def acreate(
+    max_tokens: int,
+    messages: List[Dict[str, Union[str, List[Dict[str, str]]]]],
+    model: str,
+    metadata: Optional[Dict] = None,
+    stop_sequences: Optional[List[str]] = None,
+    stream: Optional[bool] = False,
+    system: Optional[str] = None,
+    temperature: Optional[float] = 1.0,
+    thinking: Optional[Dict] = None,
+    tool_choice: Optional[Dict] = None,
+    tools: Optional[List[Dict]] = None,
+    top_k: Optional[int] = None,
+    top_p: Optional[float] = None,
+    **kwargs
+) -> Dict:
     """
-    Wrapper around Anthropic's messages API
+    Async wrapper for Anthropic's messages API
+
+    Args:
+        max_tokens (int): Maximum tokens to generate (required)
+        messages (List[Dict]): List of message objects with role and content (required)
+        model (str): Model name to use (required)
+        metadata (Dict, optional): Request metadata
+        stop_sequences (List[str], optional): Custom stop sequences
+        stream (bool, optional): Whether to stream the response
+        system (str, optional): System prompt
+        temperature (float, optional): Sampling temperature (0.0 to 1.0)
+        thinking (Dict, optional): Extended thinking configuration
+        tool_choice (Dict, optional): Tool choice configuration
+        tools (List[Dict], optional): List of tool definitions
+        top_k (int, optional): Top K sampling parameter
+        top_p (float, optional): Nucleus sampling parameter
+        **kwargs: Additional arguments
+
+    Returns:
+        Dict: Response from the API
     """
-    return await _async_anthropic_messages(*args, **kwargs)
+    return await _async_anthropic_messages(
+        max_tokens=max_tokens,
+        messages=messages,
+        model=model,
+        metadata=metadata,
+        stop_sequences=stop_sequences,
+        stream=stream,
+        system=system,
+        temperature=temperature,
+        thinking=thinking,
+        tool_choice=tool_choice,
+        tools=tools,
+        top_k=top_k,
+        top_p=top_p,
+        **kwargs,
+    )
+
+
+async def create(
+    max_tokens: int,
+    messages: List[Dict[str, Union[str, List[Dict[str, str]]]]],
+    model: str,
+    metadata: Optional[Dict] = None,
+    stop_sequences: Optional[List[str]] = None,
+    stream: Optional[bool] = False,
+    system: Optional[str] = None,
+    temperature: Optional[float] = 1.0,
+    thinking: Optional[Dict] = None,
+    tool_choice: Optional[Dict] = None,
+    tools: Optional[List[Dict]] = None,
+    top_k: Optional[int] = None,
+    top_p: Optional[float] = None,
+    **kwargs
+) -> Dict:
+    """
+    Async wrapper for Anthropic's messages API
+
+    Args:
+        max_tokens (int): Maximum tokens to generate (required)
+        messages (List[Dict]): List of message objects with role and content (required)
+        model (str): Model name to use (required)
+        metadata (Dict, optional): Request metadata
+        stop_sequences (List[str], optional): Custom stop sequences
+        stream (bool, optional): Whether to stream the response
+        system (str, optional): System prompt
+        temperature (float, optional): Sampling temperature (0.0 to 1.0)
+        thinking (Dict, optional): Extended thinking configuration
+        tool_choice (Dict, optional): Tool choice configuration
+        tools (List[Dict], optional): List of tool definitions
+        top_k (int, optional): Top K sampling parameter
+        top_p (float, optional): Nucleus sampling parameter
+        **kwargs: Additional arguments
+
+    Returns:
+        Dict: Response from the API
+    """
+    raise NotImplementedError("This function is not implemented")