aiming-lab/Agent0
A research framework for building autonomous agents that self-evolve through tool-integrated reasoning without relying on human-curated datasets.

Agent0 is a self-evolving agent framework that enables language agents to improve autonomously through multi-step co-evolution and seamless tool integration. The system demonstrates that capable agents can develop without handcrafted supervision or curated training data. The project includes Agent0 (language agents) and Agent0-VL (vision-language extension), developed by researchers from UNC-Chapel Hill, Salesforce Research, and Stanford University.
Frequently asked
- What is aiming-lab/Agent0?
- A research framework for building autonomous agents that self-evolve through tool-integrated reasoning without relying on human-curated datasets.
- Is Agent0 open source?
- Yes — aiming-lab/Agent0 is open source, released under the Apache-2.0 license.
- What language is Agent0 written in?
- aiming-lab/Agent0 is primarily written in Python.
- How popular is Agent0?
- aiming-lab/Agent0 has 1.2k stars on GitHub.
- Where can I find Agent0?
- aiming-lab/Agent0 is on GitHub at https://github.com/aiming-lab/Agent0.