GitLab Transcend - Intelligent orchestration
Mini toolsPi 0.7
π 0.7 is Physical Intelligence’s newest robotic foundation model for general-purpose embodied control. The company says it reaches specialist-level dexterous performance without task-specific fine-tuning, generalizes across robot embodiments and environments, and can recombine skills to handle novel tasks such as using unfamiliar kitchen appliances or folding laundry on a robot with no laundry-folding data. A key design point is diverse multimodal prompting: language, metadata about speed and quality, control-modality labels, and visual subgoal images, allowing the model to learn not just what to do, but how to do it.
New
Multimodal
Gen 3
Overview
π 0.7 is Physical Intelligence’s steerable general-purpose vision-language-action robot model. It is built to perform dexterous manipulation across robots, scenes, and tasks, while also showing early compositional generalization by following new language commands and combining learned skills to solve tasks not seen in training.
About Physical Intelligence
Physical Intelligence is an AI robotics company developing foundation models and learning algorithms for robots and physically-actuated devices. The company created π0 (pi-zero), the first generalist robotic foundation model capable of controlling diverse robot platforms across multiple tasks including folding laundry, assembling boxes, bussing tables, and making coffee.
Industry:
Artificial Intelligence
Company Size:
80
Location:
San Francisco, CA, US
Website:
pi.website
Tools using Pi 0.7
No tools found for this model yet.
Last updated: April 17, 2026
