3DCodeBench - Executability: leaderboard

Metric: Executability (%; Blender 5.0 scripts that run and yield a mesh, 212 instances, mean of the text-to-3D and image-to-3D tracks, single call at the best thinking level). Source: arxiv.org. Saturation forecast: Around December 2026. 12 models tracked.

Top models

#ModelScore
1Claude Opus 4.791
2GPT-5.5 (High)90.6
3GPT-5.4 (Medium)86.6
4Claude Sonnet 4.680.4
5GPT-5.4 Mini (Medium)73.1
6Gemini 3.1 Pro (Preview)72.5
7Gemma 4 31B58.2
8Gemini 3.1 Flash Lite57.5
9Gemma 4 26B51.7
10Claude Haiku 4.550.2
11Gemini 3 Flash48.1
12Gemini 3.5 Flash46.4

Interactive version: theaggregate.ai/benchmark?slug=3dcodebench-executability · How It Works · Data refreshed daily, snapshot 2026-09-26.