Xiaomi open-sources Xiaomi-Robotics-1, a VLA trained on 100K+ hours of real trajectories
Xiaomi published Xiaomi-Robotics-1 on 2026-07-16, a 5B vision-language-action model pretrained on over 100K hours of real-world UMI manipulation trajectories and post-trained on 10K+ hours of cross-embodiment data; weights (Apache-2.0) followed on Hugging Face on 2026-07-28 with top open results on RoboCasa365 and VLABench.
Key facts
- Data: 100K+ hours real-world UMI trajectories (pretraining) + 10K+ hours cross-embodiment robot data (post-training)
- RoboCasa 74.5%, RoboCasa365 57.4%, VLABench 59.1% (authors' comparison tables)
- Open weights: XiaomiRobotics/Xiaomi-Robotics-1-5B (Apache-2.0); code released 2026-08-03
- Paper reports strong scaling with data and model size
What happened
Xiaomi's robotics team released one of the largest real-data-trained open VLAs, following Xiaomi-Robotics-0 (Feb 2026).
Why it matters
It makes a 100K-hour-scale robot model openly available, narrowing the data gap between closed US labs and open Chinese releases.
Changelog
- 2026-09-29: created
Models
- Xiaomi-Robotics-1 (XR-1, 5B) Xiaomi · current
Related events
- Unitree open-sources UnifoLM-WLA-1.0 humanoid foundation model (Apache-2.0) ★★★
- Xiaomi open-sources Xiaomi-Robotics-U0, a 38B unified world model that generates multi-view robot scenes and training data ★★
Sources (3)
- paperarXiv 2607.15330: Xiaomi-Robotics-1
- codeGitHub: XiaomiRobotics/Xiaomi-Robotics-1
- codeHugging Face: Xiaomi-Robotics-1-5B
id: 2026-07-16-xiaomi-robotics-1 · updated 2026-09-29 · open in the interactive timeline