4DCodeBench Tests Coding Agents on Rebuilding Dynamic 3D Scenes from Video

Researchers from Stanford, MIT, and JHU released 4DCodeBench, a benchmark where coding agents must write executable graphics programs to reconstruct dynamic 3D scenes from video. GPT-6 Astra Max topped the leaderboard, with Claude Opus 5.5 close behind, while frontier models still struggle with dynamic physics.

2026-10-06 ~ 2026-10-06 · 4 related posts

• 4DCodeBench Shows Frontier Models Reconstruct Static Scenes but Fail at Dynamics — 4DCodeBench · 2026-10-06
• 4DCodeBench Ranks 18 Coding Agents on Inverse Graphics of Dynamic Scenes — kwangmoo_yi · 2026-10-06
• 4DCodeBench: GPT-6 Astra Max Tops 18 Agents, but More Tokens Don't Buy Better Elo — kwangmoo_yi · 2026-10-06
• 4DCodeBench: new benchmark tests coding agents on reconstructing dynamic 3D scenes from video — CSProfKGD · 2026-10-06