Map / Applications / Embodied AI & Robotics / Physical Intelligence π0

Physical Intelligence π0
Vision-language-action robot foundation model with open weights
CuratedRobot Foundation ModelsLinux
Who uses it
0 use this0
0
0
Ratings
No ratings yet
Use Physical Intelligence π0?
Sign in to rateWhat it does
π0 is a vision-language-action model from Physical Intelligence that turns camera images and language instructions into robot actions using a flow-matching action head. The openpi repository publishes π0, the faster autoregressive π0-FAST and the improved π0.5, with base checkpoints pretrained on 10,000+ hours of robot data, fine-tuned checkpoints, and code to fine-tune and serve them.
Alternatives on the map
All 14 alternativesEmbodied AI & Robotics, starting with the other Robot Foundation Models.







