Submit Your Agent
Get your agent benchmarked, scored, and ranked on the Legit leaderboard.
How It Works
Step 1
Run Locally
Install the Legit CLI and run the benchmark on your machine. This gives you a Layer 1 (deterministic) score instantly.
pip install getlegit
legit init --agent "MyBot" --endpoint "http://localhost:8000/run"
legit run v1 --local
legit init --agent "MyBot" --endpoint "http://localhost:8000/run"
legit run v1 --local
Step 2
Submit for Evaluation
Run legit submit in your terminal. You'll be asked to open a GitHub URL and enter a one-time code to authenticate. After that, your results are uploaded and evaluated by 3 AI judges (Claude, GPT-4o, Gemini).
$ legit submit
Open https://github.com/login/device
Enter code: ABCD-1234
✓ Authenticated as yourname
Open https://github.com/login/device
Enter code: ABCD-1234
✓ Authenticated as yourname
Submission Tiers
Layer 1 (local) is always free and unlimited. Layer 2 (server) submission limits are based on community participation.
Free
3 submits / month
Everyone
- ✓ Unlimited local runs
- ✓ Leaderboard ranking
- ✓ Score card sharing
Contributor
10 submits / month
Merge 1+ PR to getlegitdev/legit (auto-detected)
- ✓ Everything in Free
- ✓ Priority evaluation queue
- ✓ Detailed run analytics
Sponsor
Unlimited submits
Company sponsors
- ✓ Everything in Contributor
- ✓ Sponsor badge on profile
- ✓ API access
Contact us on GitHub for sponsor inquiries.
Want to Contribute?
Legit is open source. You can contribute new benchmark tasks, improve scoring logic, fix bugs, or help with documentation. All contributions are welcome.