🚀 NextStep-1.1 is coming！
🤖 This is a re-engineered update to the NextStep series to improve image fidelity and training stability.
🌠 Key updates:
• Higher image fidelity via extended training + Flow-based RL
• Improved numerical stability in autoregressive RL
🤗 More details are available on HF & GitHub！
HuggingFace：https://t.co/vt4yq3Ao9G
GitHub： https://t.co/DSEWk0oTt0
😎We look forward to hearing your feedback on these improvements!
open weights — A model whose trained parameters are published for anyone to download and run — unlike API-only models you can access but never possess.
Why it matters
If you work with autoregressive image models, this release targets two known weak points, artifact-prone fidelity and unstable RL training, and ships open weightsA model whose trained parameters are published for anyone to download and run — unlike API-only models you can access but never possess.Full definition → on Hugging Face and GitHub you can verify.
article👏🏻Congratulations！Step3-VL-10B was selected for HuggingFace Daily Papers…
postStep-Audio-R1.1 opens weights for a speech model that reasons in real time
