3D-DefectBench - Geometry (Silver Labels): leaderboard

Metric: Macro MCC (Matthews correlation from -1 to 1, macro-averaged over the five geometry defect types) of the model's binary per-defect verdicts against crowd majority labels on the 549-asset silver holdout; the fixed c004 design: six-view oblique RGB turntable renders of each text-to-3D asset and a rubric-guided checklist prompt. Source: arxiv.org. Saturation forecast: Around 2030. 12 models tracked.

Top models

#ModelScore
1GPT-5.40.24
2Gemini 3.1 Flash Lite0.23
3GPT-5 Mini0.21
4Gemini 2.5 Pro0.21
5Claude Opus 4.70.21
6Gemini 3.1 Pro (Preview)0.21
7Claude Sonnet 4.60.18
8GPT-4o0.17
9Qwen 3.5 397B A17B0.15
10Claude Haiku 4.50.12
11Mistral Small 3.10.09
12Qwen 2.5 VL 7B0.03

Interactive version: theaggregate.ai/benchmark?slug=3d-defectbench-geometry-silver-labels · How It Works · Data refreshed daily, snapshot 2026-09-29.