VoxelCodeBench - Symbolic Reasoning: leaderboard
VoxelCodeBench (2026) evaluates 3D spatial reasoning through code generation: from minimal API documentation and a few examples, a model writes Python that builds voxel scenes in an Unreal Engine renderer, and human annotators judge the rendered output. This board is shape correctness in percent on the 80 symbolic reasoning tasks, coordinate-based primitive placement and pattern reconstruction; higher is better.
Metric: Shape correctness (%) on the symbolic interpretation (80 tasks: coordinate-based primitive placement and pattern reconstruction) tasks, judged by a human annotator; higher is better. Source: arxiv.org. Saturation forecast: Around February 2027. 4 models tracked. Updated 7 Oct 2026.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | Claude Sonnet 4.5 | 90.3 | #261 |
| 2 | GPT-5 | 87.5 | #178 |
| 3 | Claude Opus 4 | 75 | #326 |
| 4 | Gemini 3 Pro | 15.2 | #145 |
No result here: #1 Claude Fable 5.2, #3 Gemini 4 Neon, #4 GPT-6.2 Astra.
Interactive version: theaggregate.ai/benchmark?slug=voxelcodebench-symbolic-reasoning · How It Works · Data refreshed daily, snapshot 2026-11-21.