🧠 autogen-oaiapi

OpenAI-style Chat API server for AutoGen teams.
Deploy your own /v1/chat/completions endpoint using any AutoGen-compatible team configuration.

🚀 Try it? → Don’t forget to ⭐ the repo if useful!

A self-hosted, open-source alternative to OpenAI’s ChatCompletion API, built on top of Microsoft AutoGen.

🔍 Looking for:

• OpenAI-style API server you can run locally?
• An AutoGen-based ChatCompletion implementation?
• A FastAPI wrapper for multi-agent LLM orchestration?

You found it. 🚀

WARNING

NOW(autogen-agentchat == 0.5.5) limit supported.

Because of it has bug on deserialization AssistantAgent with tools.

✨ Features

• ✅ OpenAI-compatible API interface
• ✅ Plug in any AutoGen GroupChat or SocietyOfMindAgent
• ✅ Session-aware execution (per session_id)
• ✅ FastAPI-based server with /v1/chat/completions endpoint
• ✅ stream=True response support (coming soon)

📦 Installation

pip install autogen-oaiapi

How to use?

Using just SIMPLE api!

example

client = OpenAIChatCompletionClient(
    model="claude-3-5-haiku-20241022"
)
agent1 = AssistantAgent(name="writer", model_client=client)
agent2 = AssistantAgent(name="editor", model_client=client)
team = RoundRobinGroupChat(
    participants=[agent1, agent2],
    termination_condition=TextMentionTermination("TERMINATE")
)

server = Server(team=team, source_select="writer")
server.run(host="0.0.0.0", port=8000)  # you could do not filled that args. default is that host="0.0.0.0", port=8000

Just write AutoGen team, and... Run it!

CURL call test! example

curl -X POST http://localhost:8000/v1/chat/completions \
 -H "Content-Type: application/json" \
 -d '{
 "session_id": "test-session",
 "messages": [ { "role": "user", "content": "Please write 5 funny stories." } ]
}'

Demo

Multi team support

from autogen_oaiapi.server import Server
from autogen_agentchat.agents import AssistantAgent
from autogen_agentchat.teams import RoundRobinGroupChat
from autogen_ext.models.openai import OpenAIChatCompletionClient
from autogen_agentchat.conditions import MaxMessageTermination, TextMentionTermination

client = OpenAIChatCompletionClient(
    model="gpt-4.1-nano"
)
agent1 = AssistantAgent(name="writer", model_client=client)
agent2 = AssistantAgent(name="editor", model_client=client)
team = RoundRobinGroupChat(
    participants=[agent1, agent2],
    termination_condition=TextMentionTermination("TERMINATE"),
)

server = Server()
server.model.register(
    name="TEST_TEAM",
    actor=team,
    source_select="writer",
)
server.run(port=8001)

or

from autogen_oaiapi.server import Server
from autogen_agentchat.agents import AssistantAgent
from autogen_agentchat.teams import RoundRobinGroupChat
from autogen_ext.models.openai import OpenAIChatCompletionClient
from autogen_agentchat.conditions import MaxMessageTermination, TextMentionTermination

server = Server()

@server.model.register(name="TEST_TEAM_DECORATOR", source_select="writer")
def build_team():
    client = OpenAIChatCompletionClient(
        model="gpt-4.1-nano"
    )
    agent1 = AssistantAgent(name="writer", model_client=client)
    agent2 = AssistantAgent(name="editor", model_client=client)
    team = RoundRobinGroupChat(
        participants=[agent1, agent2],
        termination_condition=TextMentionTermination("TERMINATE"),
    )
    return team

server.run(port=8001)

Look at the example folder include more examples!

• simmple example
• function style register example
• decorator sytle register example
• JSON file style api_key register example
• Memory style api_key register example

Star History