CADEngBench - Assembly End-to-End (A2): leaderboard
Metric: End-to-end pass rate (%; share of 120 assembly relation tasks between two source B-Rep bodies; one response per task whose rank-1 joint is correct and whose predicted joint frame, placement and allowed and blocked motions pass the scripted kinematic tests). Source: arxiv.org. Saturation forecast: Around 2030. 8 models tracked.
Top models
| # | Model | Score |
|---|---|---|
| 1 | Gemini 3 Flash (Preview) (Minimal) | 15.8 |
| 2 | Claude Sonnet 4.5 | 12.5 |
| 3 | GPT-5.2 (Non-reasoning) | 11.7 |
| 4 | Llama 4 Maverick | 10.8 |
| 5 | Qwen 3.5 35B A3B (Non-reasoning) | 10.8 |
| 6 | GLM-4.6V (Non-reasoning) | 8.3 |
| 7 | Kimi K2.5 (Non-reasoning) | 7.5 |
Interactive version: theaggregate.ai/benchmark?slug=cadengbench-assembly-end-to-end-a2 · How It Works · Data refreshed daily, snapshot 2026-09-29.