multimodal benchmark
BenchCAD Leaderboard
BenchCAD is a multimodal benchmark for programmatic CAD reasoning built from 17,900 execution-verified CadQuery programs spanning 106 industrial part families.
Updated Sep 5, 2026
Models4
Model coverage4
MetricScore
EvidenceB
BenchCAD Ranking
Higher score ranks better on this benchmark.
Sort by
Rank | Model | Score | Percentile | Participants | Evidence | Evaluated |
|---|
| Rank01 | ModelOP | Score70.60% | Percentile100.00% | Participants4 | EvidenceC | Evaluated |
| Rank02 | ModelOP | Score63.10% | Percentile66.67% | Participants4 | EvidenceC | Evaluated |
| Rank03 | ModelOP | Score62.30% | Percentile33.33% | Participants4 | EvidenceC | Evaluated |
| Rank04 | ModelAN | Score37.30% | Percentile0.00% | Participants4 | EvidenceC | Evaluated |
BenchCAD Highlights
The leading models and scores on this benchmark.
BenchCAD Score Distribution
A closer view of the leading scores on this benchmark.
The Top AI Models for BenchCAD
The first five results on this benchmark, with official price and output speed added where the model identity can be matched.
What is BenchCAD?
What BenchCAD measures and how its scores work.
BenchCAD is a benchmark that decomposes CAD capability into matched tasks, including Vision2Code, which requires models to generate CadQuery code from multi-view renders.
The Vision2Code task is scored by voxel IoU against the reference geometry, reported as a ratio.
Scores are shown in ratio. This benchmark is not independently verified and has an evidence level of B.
- Family
- BenchCAD
- Modality
- multimodal
- Primary category
- multimodal
- Score direction
- higher
- LLMBoard eligible
- No
- Evaluation key
- overall
LLM Stats. Benchmark scores retain their original unit. Overall score eligibility is shown separately.
FAQ
Common questions about BenchCAD.
Which model scores highest on BenchCAD?
GPT-5.6 Sol is currently ranked first with 70.60%.
What are the top three models on BenchCAD?
The current leaders are GPT-5.6 Sol (70.60%), GPT-5.6 Luna (63.10%), and GPT-5.6 Terra (62.30%).
Which BenchCAD model has the lowest official input price?
GPT-5.6 Luna has the lowest matched official input price at $0.20 input / $1.2 output per 1M tokens.
Which models are fastest among BenchCAD results?
The fastest matched records are GPT-5.6 Terra (99.35 tok/s via OpenAI), Claude Sonnet 5 (42.00 tok/s via Anthropic), and GPT-5.6 Luna (41.87 tok/s via OpenAI).
Does the highest score result prove overall model quality?
No. This benchmark measures one defined capability or task. The overall LLMBoard score uses a separate aggregation across eligible benchmark evidence.