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

Metric: Macro MCC (Matthews correlation from -1 to 1, macro-averaged over the four texture 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 2032. 12 models tracked.

Top models

#ModelScore
1Gemini 3.1 Pro (Preview)0.16
2GPT-5.40.15
3Gemini 3.1 Flash Lite0.15
4Gemini 2.5 Pro0.12
5Claude Opus 4.70.11
6GPT-4o0.1
7GPT-5 Mini0.07
8Mistral Small 3.10.06
9Qwen 3.5 397B A17B0.05
10Claude Sonnet 4.60.03
11Qwen 2.5 VL 7B0.02
12Claude Haiku 4.50.02

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