Skip to content

BenchCAD Vision2Code voxel IoU without tools

Generates CadQuery code from multi-view renders and scores geometric similarity by voxel intersection-over-union.

3 models

Higher is better.

  1. 1Claude Sonnet 5.5Anthropic0.747
  2. 2Claude Opus 5.5Anthropic0.730
  3. 3Claude Opus 5Anthropic0.366
Higher is better. Bar colour: blue proprietary, orange open weights, green other.

About this benchmark

Metric
Voxel IoU
Tasks
1,000-file Vision2Code subset
Authors
Zhang et al. and Anthropic (2026)

Scores are as reported by model developers or evaluators and may use different settings (for example reasoning effort or tools). Compare with care.