AutoGen 0.7.5: our test log

AutoGen 0.7.5 passed our one-tool task on a local 3B model: 3 of 3 trials; cold uv install 1.9 s.

Result at a glance

What we found

• 3 of 3 task trials passed: the model's tool call reached the framework, the function ran with 17 and 25, and the final answer contained 42.
• Cold install took 1.9 s with uv and 4.8 s with pip, and left 59 MB in 28 packages (uv).
• Used the 0.7 AgentChat API (AssistantAgent with a model client from autogen-ext). Code written for the older 0.2 API will not run against this package.

What we ran, step by step

# Install with uv (cold, no cache)
uv venv -q -p 3.13 work/py/autogen && VIRTUAL_ENV=work/py/autogen uv pip install --no-cache -q autogen-agentchat "autogen-ext[openai]"

# Install with pip (cold, no cache)
uv venv -q --seed -p 3.13 work/pip/autogen && work/pip/autogen/bin/pip install --no-cache-dir -q autogen-agentchat "autogen-ext[openai]"

# Task command (all trials)
OTEL_SDK_DISABLED=true work/py/autogen/bin/python tasks/autogen_agent.py

Raw output (excerpts)

from common import *
import asyncio
from autogen_agentchat.agents import AssistantAgent
from autogen_ext.models.openai import OpenAIChatCompletionClient
client = OpenAIChatCompletionClient(model=MODEL, base_url=BASE, api_key=KEY, temperature=0, model_info={"vision":False,"function_calling":True,"json_output":False,"family":"unknown","structured_output":False})
agent = AssistantAgent("assistant", model_client=client, tools=[add_numbers], reflect_on_tool_use=True)
async def main():
    r = await agent.run(task=TASK)
    await client.close()
    return r.messages[-1].content
report(asyncio.run(main()))

import json, os
BASE = "http://127.0.0.1:11434/v1"
MODEL = os.environ.get("MODEL", "qwen2.5-3b")
KEY = "ollama"
TASK = "Use the add_numbers tool to compute 17 + 25, then tell me the result."
CALLS = []
def add_numbers(a: int, b: int) -> int:
    """Add two integers and return the sum."""
    CALLS.append((a, b))
    return int(a) + int(b)
def report(final):
    ok = any({int(x), int(y)} == {17, 25} for x, y in CALLS) and "42" in str(final)
    print("RESULT_JSON " + json.dumps({"calls": CALLS, "final": str(final)[:300], "ok": ok}))

RESULT_JSON {"calls": [[17, 25]], "final": "The result of adding 17 and 25 is 42.", "ok": true}

Gotchas and things that broke

• Without model_info the client raises ValueError: model_info is required when model name is not a valid OpenAI model; we reproduced that error and then set model_info with function_calling true.

What we did not test

• Any vendor-hosted model: everything ran on a local 3B model, so tool-call reliability with frontier models is not measured here.
• Multi-agent flows, memory, long conversations, streaming, tracing and cloud deployment.
• Anything beyond one tool and one question; three trials is a smoke test, not a benchmark.

Reproduce it

Versions change. These commands install whatever is latest on the day you run them unless a version is pinned, so expect different results later. The version we measured is in the table above.

# 1. Serve any local model that supports tool calling behind an OpenAI-compatible endpoint, registered as qwen2.5-3b:
#    http://127.0.0.1:11434/v1  (we used Ollama with a Qwen2.5-3B-Instruct Q4_K_M GGUF)
uv venv -p 3.13 .venv && VIRTUAL_ENV=.venv uv pip install autogen-agentchat "autogen-ext[openai]"
# 2. Save common.py (above) and the task script next to each other, then:
.venv/bin/python autogen_agent.py

How to read this test

Every number on this page was measured on this one machine, which also ran other jobs, so timings are indicative and vary run to run. A pass means our small task worked once or as stated, a fail means it did not work as we ran it; neither says how the tool behaves with a vendor's frontier model, on a GPU, or at scale. Where a tool could not run without an account or a GPU, we say so instead of guessing.

See the testing rig for the hardware, the local model and the rules we follow, and the test hub for every other test.

Related tests: agent frameworks and sdks

• CrewAI: CrewAI 1.15.30 passed our one-tool task on a local 3B model: 3 of 3 trials; cold uv install 8.4 s.
• DSPy: DSPy 3.4.1 passed our one-tool task on a local 3B model: 3 of 3 trials; cold uv install 3.7 s.
• Google ADK (Python): Google ADK (Python) 2.13.0 passed our one-tool task on a local 3B model: 3 of 3 trials; cold uv install 1.7 s.
• Haystack: Haystack 3.4.1 passed our one-tool task on a local 3B model: 3 of 3 trials; cold uv install 2.3 s.
• Instructor: Instructor 1.17.0 returned a validated Pydantic object from a local 3B model in 3 of 3 trials (structured extraction, not a tool call); cold uv install 1.4 s.
• LangChain: LangChain 1.4.7 passed our one-tool task on a local 3B model: 3 of 3 trials; cold uv install 0.8 s.

All test logs · How we test

FAQ

Sources

• AutoGen repository (fetched 2026-10-25)

Related

• All test logs
• Methodology and testing rig
• Study: install size and first-task results across 16 frameworks
• AutoGen facts and sources