π₀ / openpi
Physical Intelligence
A foundation-policy reference for studying adaptation and continuous action chunks.
What you can explore
Images, robot state and task prompt, through a platform-specific adapter. Continuous action chunks for π₀; do not conflate them with π₀-FAST tokenization.
Before you use it
The release includes ALOHA and DROID configurations. The maintainers explicitly caution that checkpoints may not transfer to another setup.
Original contributors
Physical Intelligence
License / access: Repository code lists Apache-2.0; inspect checkpoint and bundled base-model terms separately, including LICENSE_GEMMA.
Keep exploring
DROID
Robot demonstrations collected in varied scenes using a shared Franka manipulation setup.
ACT / Action Chunking with Transformers
An imitation-learning implementation that predicts chunks of robot actions from demonstrations.
LeRobot
Tools for recording datasets, training policies and exploring robot-learning workflows.
Editorial reference · checked 2026-09-25