MindCube
Higher is better · Image-input models only
Multiple-choice spatial reasoning from a few photos of a real scene taken from different viewpoints: where things are, which way they face, how the scene would look after moving, and what hidden parts are like. Humans get about 95% right. Scores run from 0 to 100%. Higher is better.
Top score44.8%Claude Sonnet 4
Models tested1
Model release date (newest on the right)
Best score so farHigher on the chart is better
Which AI gives the photo and chart answers people prefer most?Arena Vision
Claude Fable 5🥇Claude Fable 5🥇Can it spot the wrong step from assembly photos?Furniture AssemblyClaude Opus 5.5🥇Claude Opus 5.5🥇Can it draw a floor plan from a few photos of a home?Blueprint-Bench 2
Gemini 4 Argon🥇Gemini 4 Argon🥇
Claude Fable 5🥇Claude Fable 5🥇Can it spot the wrong step from assembly photos?Furniture AssemblyClaude Opus 5.5🥇Claude Opus 5.5🥇Can it draw a floor plan from a few photos of a home?Blueprint-Bench 2Gemini 4 Argon🥇Gemini 4 Argon🥇- 🥇AnthropicClaude Sonnet 444.8%
Last updated 2026-10-07