Model quality leaderboard
Every score comes from the same evaluation pipeline — reproducible and traceable. The leaderboard updates as new results arrive.
Leaderboard dimension
Overall ranking
| Rank | Model | Overall score | Grade | Evaluations |
|---|---|---|---|---|
| 1 | pm2 | 8.3 | D | 3 |
| 2 | GPT-6 Astra | 0.0 | D | 11 |
| 3 | pm1 | 0.0 | D | 1 |