Video-Coder: first code-driven video generation agent trained with execution rewards
UCF-Ai-LLMs · hf · 2026-11-09
Researchers from UCF, Microsoft Research, Tsinghua, and UNC introduce Video-Coder, the first code-driven video generation agent: it autonomously writes and executes code to construct dynamic 3D scenes (DRAFT Video), providing structure and motion signals to downstream video generators for precise spatial layout, coherent motion, and controllable composition.
The team built Video-Coder Harness (a unified environment integrating code execution, 3D rendering, and execution feedback) plus two curated datasets (SFT-6k, RL-2k), training via SFT followed by RL with dual execution- and video-based rewards.
More from coding & agent
- One-person site with 27 followers made $4,800 in a day, found via Claude Code revenue scan — FinanceYF5 · 2026-11-09
- The churn era: customers now use Claude to slop up replacements for overpriced SaaS — joshgonsalves_ · 2026-11-09
- Designer builds AI pipeline that finally nails map interchange generation after 2 weeks — tomjohndesign · 2026-11-09
- AI lets designers build wild ideas, but knowing the low-level stack matters more than ever — nptacek · 2026-11-09
- Gemini CLI PR fixes orphaned child processes and wrong exit codes on headless abort — duvanka-prasad · 2026-11-09
- T3 Code hands-on: one GUI orchestrates Claude Code and Codex sessions — vista8 · 2026-11-09