4DCodeBench: leaderboard
Agentic reconstruction of 200 videos (100 real and 100 simulated) as executable graphics programs. The score is the equal-weight mean of five reconstruction families on fixed [0,1] scales: perception, 2D motion, depth, 3D shape and 3D motion. Failed renders incur declared penalties. Models use the published provider-specific coding harnesses and effort settings; VLM preference Elo and VQA are separate measurements.
Metric: Reconstruction score (0–1), five-family mean over 200 scenes; provider-specific coding harnesses. Source: 4dcodebench.com. Saturation forecast: Around February 2027. 16 models tracked. Updated 7 Oct 2026.
Top models
| # | Model | Setting | Score | Overall rank |
|---|---|---|---|---|
| 1 | GPT-6 Astra | Max | 0.7908 | #11 |
| 2 | Claude Opus 5.5 | High | 0.7784 | #9 |
| 3 | Claude Fable 5.1 | High | 0.741 | #20 |
| 4 | Claude Opus 5 | High | 0.7172 | #26 |
| 5 | GPT-5.6 Luna | Max | 0.6695 | #135 |
| 6 | GPT-5.6 Sol | High | 0.6454 | #36 |
| 7 | Gemini 3.8 Flash | High | 0.6394 | #43 |
| 8 | DeepSeek V4.1 Flash | High | 0.5965 | #69 |
| 9 | Qwen 3.8 Flash | xHigh | 0.5627 | #132 |
| 10 | GPT-5.6 Terra | High | 0.5505 | #74 |
| 11 | GLM-5.3 Flash | Max | 0.4732 | #137 |
| 12 | Muse Glimmer 30B | High | 0.4419 | #395 |
| 13 | MiMo-V2.5 | 0.4348 | #387 | |
| 14 | MiniMax-M3 | 0.3394 | #290 | |
| 15 | Mistral Medium 3.5 | 0.3152 | #742 |
No result here: #2 Claude Fable 5.2, #3 Gemini 4 Neon, #4 GPT-6.2 Astra.
Interactive version: theaggregate.ai/benchmark?slug=4dcodebench · How It Works · Data refreshed daily, snapshot 2026-11-22.