Benchmarks · reasoning
EnigmaEval
1,184 long-form multimodal puzzles from puzzle hunts such as the MIT Mystery Hunt, graded by exact string match; accuracy over three runs. Private test set (Scale Labs, CAIS, MIT).
5 models · 5 results · data as of 2026-10-29 · labs.scale.com ↗
Leaderboard
Fable 5 high
All results
5 rows| # | Model | Effort | EnigmaEval |
|---|---|---|---|
| 1 | Fable 5 | high | 39.3% ±2.8 |
| 2 | GPT-5.6 Sol | high | 37.1% ±2.8 |
| 3 | Gemini 3.1 Pro | high | 36.8% ±2.7 |
| 4 | Gemini 3.5 Flash | high | 25.4% ±2.4 |
| 5 | Opus 4.8 | xhigh | 23.5% ±2.4 |