WRSWorld Research StateContribute
WORLD RESEARCH STATE

Open-Sora2: hierarchical data and efficient open video-model training system

Contribution grade

Open video-generation systems · Human-designed research method or resource

Reference score60,000Provisional editorial estimate of this exact contribution; its grade is assessed separately.

What changed

Integrated hierarchical filtering/annotation, progressive training, VideoDC-AE/DiT architecture and system optimizations in an openly released video-generation training system.

Scientific scope

Integrated hierarchical filtering/annotation, progressive training, VideoDC-AE/DiT architecture and system optimizations in an openly released video-generation training system. Inherits diffusion transformers, compression and video/data-processing methods. Reported dollar budget is source-defined rather than total historical R&D/data cost or a current reproducible quote. Human preference tests use100prompts, not universal commercial equivalence. Distinct from OpenAI Sora and PKU Open-Sora-Plan.

Evidence

The linked original sources describe the stated contribution and its evidence limits. WRS has not independently rerun the experiments or certified a proof. Inherits diffusion transformers, compression and video/data-processing methods. Reported dollar budget is source-defined rather than total historical R&D/data cost or a current reproducible quote. Human preference tests use100prompts, not universal commercial equivalence. Distinct from OpenAI Sora and PKU Open-Sora-Plan.

Contributors

Provisional WRS assessment, edition 2026-10-16.1. Independent expert calibration is pending. This grade does not certify correctness or assign ownership.

Full scientific record · Browse contribution grades · Request a correction