benchgap
Vision & documents

BenchCAD Vision2Code (tools) leaderboard

As of 2026-11-19, the highest measured score on BenchCAD Vision2Code (tools) is 1.0% by Claude Sonnet 5.5. 4 more models have estimated scores, calibrated from the benchmarks they were measured on.

Measured scores: benchlm.ai.

#ModelScoreSource
1Claude Fable 5.21.0%estimated ± 0.0 pp, low confidence
2Claude Sonnet 5.51.0%measured
3Claude Opus 5.51.0%measured
4GPT-6 Astra1.0%measured
5Kimi K3.1 Pro0.9%measured
6Claude Fable 5.10.8%measured
7Claude Opus 50.8%measured
8DeepSeek V4.1 Flash 10170.8%estimated ± 0.0 pp, medium confidence
9Inkling 2 Small0.8%estimated ± 0.0 pp, low confidence
10GLM-5.3-Flash0.7%estimated ± 0.0 pp, low confidence

All leaderboards

Agentic · terminal

Agentic · tools

Coding

Math

Knowledge & reasoning

Instruction following

Multilingual

Vision & documents

Long context

General