Xiaomi has introduced Xiaomi-Robotics-U0, a 38-billion-parameter multimodal autoregressive model designed for unified embodied synthesis. The model treats embodied generation as an extension of foundation image and video generation, jointly optimizing text-to-image generation, image editing, embodied scene generation, embodied transfer, and embodied video generation.
- Preserves the generalization of pre-trained world foundation models while adapting them to embodied settings.
- Supports high-quality multi-view scene generation across multiple robot embodiments.
- Introduces structured, controllable embodied transfer for fine-grained editing while preserving multi-view consistency.
- Outperforms GPT-Image-2.0 in human evaluations of embodied scene generation and transfer.
- Ranks first on World Arena for embodied video generation.
- Improves the out-of-distribution success rate of pi_0.5 from 36.9% to 63.2% on challenging real-world manipulation tasks.
The results demonstrate that foundation world models can serve as both embodied world models and scalable data engines for embodied intelligence.