Oriel — AI & technology worth noticing, every day.

Video-Coder: Reinforcing Agentic Coding for Video Generation

Paper · https://arxiv.org/abs/2611.09894

Recent years have witnessed remarkable progress in video generation in terms of visual quality and temporal consistency. However, generating videos with precise spatial structure, coherent motion dynamics, and controllable scene composition remains challenging. In this paper, we introduce Video-Coder, the first attempt to train a code-driven video generation agent, which autonomously writes and executes code to construct a dynamic 3D scene (Draft Video) that provides structure and motion signals to downstream video generators. To support this paradigm, we design a dedicated data pipeline and curate two high-quality datasets, Video-Coder-SFT-6k and Video-Coder-RL-2k, which contain diverse prompts and ground-truth reference videos. We further develop the Video- Coder Harness, a unified interactive environment integrating code execution, 3D scene rendering, and execution feedback for agent training and inference. We then train Video-Coder with SFT fol- lowed by harness-native agentic reinforcement learning with dual reward feedback, which combines execution-based and video-based rewards. Additionally, we design an annealing training strategy that gradually reduces the execution reward over time, allowing the model to rely more on the video reward in the later training stages. Experiments show that Video-Coder achieves significant gains, improving Minimax-H3 by approximately 7 points on PhyGenBench and 24 points on VBench-2.0. We hope this work can serve as an open foundation for coding agents in video generation, and we fully open-source our data, models, and code

First seen by Oriel on .

AI summary

Video-Coder is a code-driven video generation agent that builds dynamic 3D draft scenes, improving spatial structure, motion control, and downstream generation benchmark scores.

  • The agent autonomously writes and executes code to produce Draft Videos containing structural and motion signals for downstream video generators.
  • Video-Coder Harness combines code execution, 3D scene rendering, and execution feedback in one environment for training and inference.
  • Training uses Video-Coder-SFT-6k and Video-Coder-RL-2k, followed by reinforcement learning with execution-based and video-based rewards.
  • An annealing strategy gradually reduces execution reward, while Video-Coder improves Minimax-H3 by approximately 7 points on PhyGenBench and 24 points on VBench-2.0.
  • The authors fully open-source the project’s data, models, and code as a foundation for coding agents in video generation.

#video generation #coding agents #reinforcement learning #3d rendering

Generated by gpt-5.6-sol. Check the original for details.

Metric trends

Last 30 days · Up to 120 recorded samples per source and metric, with reposts kept separate · First and latest retained

Points

Score reported by the source.

  • Hugging Face Papers: 22, latest observation ; +11 over observed period since

Comments

Comment count reported by the source.

  • Hugging Face Papers: 3, latest observation ; +2 over observed period since

Aliases

SourceExternal IDURL
Hugging Face Papers2611.09894https://arxiv.org/abs/2611.09894 Read discussion