InsanityBench: leaderboard
10 handcrafted cryptic puzzles mixing stories, images, and cryptic number sequences that probe lateral thinking and creative leaps. Tasks require visual interpretation (e.g. recognizing objects in images) combined with wordplay and cross-domain reasoning.
Metric: Total Score. Source: robinhaselhorst.com. Source: robinhaselhorst.com. Saturation forecast: Around 2029. 8 models tracked.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | Gemini 3.1 Pro (Preview) | 15 | #58 |
| 2 | GPT-5.4 (High) | 10 | #79 (GPT-5.4) |
| 3 | GPT-5.2 (High) | 10 | #110 (GPT-5.2) |
| 4 | Claude Opus 4.6 (Medium) | 10 | #63 (Claude Opus 4.6) |
| 5 | Grok 4 | 5 | #181 |
| 6 | Gemini 3 Flash (High) | 5 | #99 (Gemini 3 Flash) |
| 7 | Grok 4.1 Fast (Reasoning) | 0 | #217 (Grok 4.1 Fast) |
| 8 | GPT-5.2 (Low) | 0 | #110 (GPT-5.2) |
No result here: #3 Claude Opus 5.5, #6 GPT-6 Astra, #7 GPT-6.1 Sol.
Interactive version: theaggregate.ai/benchmark?slug=insanitybench · How It Works · Data refreshed daily, snapshot 2026-10-20.