Keeping up with AI open source is a full-time job. Every week a new agent framework, MCP server or LLM tool shoots up GitHub. I got tired of jumping between news sites, GitHub trending and X threads, so I built a small site called AIBARS that tracks it all in one place.
Here is a snapshot of what its rankings show today (Sep 30, 2026).
Most starred AI projects
- openclaw: ~390k stars
- superpowers: ~292k stars
- skills: ~270k stars
- hermes-agent: ~249k stars
- deepseek-harness: ~237k stars
- n8n: ~206k stars
Fastest star growth
- deepseek-harness: +3,408
- skills: +2,198
- ponytail: +1,906
- cc-switch: +1,871
- firecrawl: +1,678
What I noticed
- Agent tooling dominates the charts. Agent harnesses, skills and workflow projects fill most of both lists. The attention has clearly moved from models to the tooling around them.
- Total stars and star growth tell different stories. Some projects with huge star counts are barely moving, while smaller ones like ponytail and cc-switch are climbing fast. If you only look at the all-time list, you miss what is rising right now.
- The same names show up twice. deepseek-harness and skills appear in both the most-starred and fastest-growing lists, which suggests momentum is compounding for a few winners.
Why I built it
I wanted a single page that answers "what actually matters in AI this week?" without opening ten tabs. AIBARS started as my own dashboard for that, and I've kept improving it for over a year.
Besides the open-source rankings, it has daily curated AI news, model rankings, 100+ AI tools and free learning resources. It is available in 6 languages: English, Simplified Chinese, Traditional Chinese, Japanese, Korean and German.
You can see the live rankings here: https://aibars.net/en/ranking
Your turn
What do you use to keep up with AI open source? And what would you want a site like this to show that it doesn't yet?
Top comments (1)
One thing I’ve noticed while tracking fast-growing open-source AI projects is that GitHub stars alone don’t tell the whole story. I look at release frequency, contributor growth, community activity, documentation, and how quickly projects start appearing in real-world workflows.
I also like checking how these projects and brands show up across search and AI-generated results. SerpSpur has been useful for that side of the research, especially for looking at AI mentions, citations, and broader visibility.
The interesting part is watching a project go from “interesting GitHub repo” → community adoption → ecosystem → actual search/AI visibility. That transition often tells you more than a star count.