Vending Bench 2 benchmark scores from Epoch AI's benchmark data collection.
This definition has not completed editorial and provenance review. Verify its metric, version, source, and evaluation configuration upstream before using it for a model-selection decision.
Sourced from Epoch AI's systematic benchmark data collection covering frontier models.
Auto-generated definition — editorial description not yet written.
Higher is better.
Auto-generated definition — editorial description not yet written.
| # | Model | Score |
|---|---|---|
| 1 | 10936.8 | |
| 2 | 6204.6 | |
| 3 | 5940.1 | |
| 4 | 5634.4 | |
| 5 | 5478.2 | |
| 6 | 5114.9 | |
| 7 | 5082.9 | |
| 8 | 4967.1 | |
| 9 | 4967.1 | |
| 10 | 4662.8 | |
| 11 | Claude Fable 5 (Adaptive Reasoning, Max Effort, Opus 4.8 Fallback) source epoch_ai Evidence observed Jul 31, 2026 | 4529.9 |
| 12 | 4432.1 | |
| 13 | 4254.2 | |
| 14 | 3887.4 | |
| 15 | 3838.7 | |
| 16 | 3838.7 | |
| 17 | 3634.7 | |
| 18 | 3634.7 | |
| 19 | 2992.3 | |
| 20 | 2376.8 | |
| 21 | 2157.8 | |
| 22 | 1198.5 | |
| 23 | 1106.6 | |
| 24 | 911.2 | |
| 25 | 573.6 | |
| 26 | 548.8 | |
| 27 | 462.7 | |
| 28 | 458.9 | |
| 29 | 458.9 | |
| 30 | 202.0 | |
| 31 | 160.6 | |
| 32 | 71.6 | |
| 33 | 35.3 | |
| 34 | 0.5 | |
| 35 | -11.3 | |
| 36 | -21.5 | |
| 37 | -23.2 |