EnigmaEval

Puzzle-hunt style multi-step reasoning. Scale AI.

Published by

Scale AI SEAL

Category

Reasoning

Index weight

0.5

Models

5

Data as of

9 Sept 2026

Scale AI SEAL Leaderboards.

1. 1Claude Fable 539.3%
2. 2GPT-5.6 Sol37.1%
3. 3Gemini 3.1 Pro36.8%
4. 4Gemini 3.5 Flash25.4%
5. 5Claude Opus 4.823.5%

SearchAll providersAnthropicGoogleOpenAIOpen weights5 of 5

#	Model	Score ↓	Run ↕	Index ↕	Published ↕
1	Claude Fable 5Anthropic	39.3%	–	63.9	–
2	GPT-5.6 SolOpenAI	37.1%	–	68.0	–
3	Gemini 3.1 ProGoogle	36.8%	–	60.7	–
4	Gemini 3.5 FlashGoogle	25.4%	–	60.4	–
5	Claude Opus 4.8Anthropic	23.5%	–	–	–