Following my post on open-source local LLM inferencing options, I wanted to perform a similar exercise for open-source LLM agentic frameworks. So I could see which projects had the largest community backing and mature development before selecting a framework for my own projects.
For optimal viewing experience, please use a desktop browser. Mobile users may find some columns hidden due to screen size limitations.
The table above gives an in-depth look at popular open-source agentic frameworks, organised by star count (a common indicator of community interest). The categories used here are subjective and can be interpreted in various ways.
For this analysis, I’ve used a broad definition of AI agents to cover a wide range of projects. This includes comprehensive agentic frameworks like crewAI, as well as coding assistant plugins like Continue. Each project reflects different aspects of AI agency, whether in part or in full.
If you have suggestions or know of other projects I missed, feel free to open an issue on my GitHub repo.
You can read about AI Agents and Agentic Workflows here.