Skip to main content

​
Overview

AutoGen can use TokenLab through OpenAIChatCompletionClient, the same OpenAI-compatible client path it uses for other custom Chat Completions endpoints.
Type: Agent FrameworkPrimary Path: OpenAI-compatible chat completionsSupport Confidence: Supported custom endpoint path

​
Installation

pip install -U autogen-agentchat "autogen-ext[openai]" pyyaml

​
Environment

export TOKENLAB_API_KEY="sk-your-tokenlab-key"

​
YAML Model Config

Use TokenLab’s /v1 endpoint as base_url: 
provider: autogen_ext.models.openai.OpenAIChatCompletionClient
config:
  model: claude-sonnet-5
  base_url: https://api.tokenlab.sh/v1
  api_key: ${TOKENLAB_API_KEY}
  model_info:
    function_calling: true
    json_output: true
    vision: false
    family: claude

​
Python Client

import os

from autogen_agentchat.agents import AssistantAgent
from autogen_ext.models.openai import OpenAIChatCompletionClient

model_client = OpenAIChatCompletionClient(
    model="claude-sonnet-5",
    base_url="https://api.tokenlab.sh/v1",
    api_key=os.environ["TOKENLAB_API_KEY"],
    model_info={
        "function_calling": True,
        "json_output": True,
        "vision": False,
        "family": "claude",
    },
)

agent = AssistantAgent("assistant", model_client=model_client)

​
Endpoint Notes

AutoGen’s OpenAI extension expects the OpenAI-compatible request shape. For provider-native features such as Anthropic Messages or Gemini generateContent, use TokenLab’s native endpoints from a client that supports those protocols directly.