OpenAI GPT-5.6 Launch - BenchCAD: leaderboard
BenchCAD row of the computer use table in OpenAI's GPT-5.6 launch post (2026-07-09): models reconstruct 3D objects from multi-view renders by writing CAD code, without the python tool reported on a separate row. Score in percent, higher is better; includes Claude Mythos 5, Mythos Preview and Opus 4.8.
Metric: Score (%). Source: openai.com. Saturation forecast: Not forecast. 7 models tracked.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | GPT-5.6 Sol | 70.6 | #22 |
| 2 | GPT-5.6 Luna | 63.1 | #79 |
| 3 | GPT-5.6 Terra | 62.3 | #40 |
| 4 | GPT-5.5 | 44.4 | #33 |
| 5 | Claude Mythos 5 | 38.4 | #8 |
| 6 | Claude Mythos Preview | 35.5 | #5 |
| 7 | Claude Opus 4.8 | 27.3 | #39 |
No result here: #3 Claude Opus 5.5, #6 GPT-6 Astra, #7 GPT-6.1 Sol.
Interactive version: theaggregate.ai/benchmark?slug=openai-gpt-5-6-launch-benchcad · How It Works · Data refreshed daily, snapshot 2026-11-05.