Brain
Physical Intelligence's open VLA flow model. A roughly 3B-parameter VLM backbone produces motor commands at up to 50 Hz across multiple robot platforms. Trained on internet-scale vision-language data, Open X-Embodiment, and Physical Intelligence's dexterous-manipulation dataset. The first robotics foundation model ported to Hugging Face LeRobot. pi0.5 (September 2025) added open-world generalization.
Foundation model · Maturity: Research · Open source
Machine-readable surfaces
- Markdown mirror: /brains/pi0.md
- RSS feed: /brains/pi0/feed.xml
- JSON-LD: embedded in this page’s head
- REST API: /v1/brains/723c2abe-101d-4132-bbfb-663c1f517d3e
- Data documentation: /data
Architecture
VLA flow model, roughly 3B-parameter VLM backbone, motor commands at up to 50 Hz, cross-embodiment across 8 platforms: UR5e, Bimanual UR5e, Franka, Bimanual Trossen, Bimanual ARX, Mobile Trossen, Mobile Fibocom. Trained on internet-scale vision-language data, Open X-Embodiment, and Physical Intelligence's dexterous-manipulation dataset. PyTorch and JAX implementations.
Key facts
- DEPLOY maturity rationale
- Open foundation model in broad developer use; re-derived as research-to-commercial; recorded here as research given primary use is research/development
- Open-sourced
- February 2025 via the openpi repository
- Tuning data efficiency
- 1 to 20 hours of data sufficient to tune for a new platform
- pi0.5
- September 2025: added open-world generalization
- Powered platforms (research, not registry models)
- UR5e, Bimanual UR5e, Franka, Bimanual Trossen, Bimanual ARX, Mobile Trossen, Mobile Fibocom
- Hosted by
- Hugging Face LeRobot
Developed by (1)
Hosted by (1)
- LeRobot(Framework)
Sources (3)
- Physical Intelligence: openpi repository (primary) · https://github.com/Physical-Intelligence/openpi
- Hugging Face Blog: pi0 on LeRobot · https://huggingface.co/blog/pi0
- The Robot Report: pi0 coverage · https://www.therobotreport.com/
Canonical ID 723c2abe-101d-4132-bbfb-663c1f517d3e