KnotBench - Same Knot (Images): leaderboard
The A0-I task of KnotBench (2026, The Gordian Knot for VLMs): judge whether two rendered diagrams show the same knot, with answers checked against Regina canonical knot signatures. The score is accuracy in percent under strict answer matching, with a random baseline of 50 percent; higher is better.
Metric: Accuracy (%) on the 200 A0-I items: whether two rendered diagrams show the same knot; zero-shot with a forced ANSWER line and strict matching, 64K output-token budget, random baseline 50%; higher is better. Source: arxiv.org. Saturation forecast: Around February 2027. 2 models tracked. Updated 7 Oct 2026.
Top models
| # | Model | Setting | Score | Overall rank |
|---|---|---|---|---|
| 1 | Claude Opus 4.7 | Thinking | 55 | #85 |
| 2 | GPT-5 | Thinking | 50 | #178 |
No result here: #1 Claude Fable 5.2, #3 Gemini 4 Neon, #4 GPT-6.2 Astra.
Interactive version: theaggregate.ai/benchmark?slug=knotbench-same-knot-images · How It Works · Data refreshed daily, snapshot 2026-11-21.