Physical Intelligence π0.5
A generalization-focused evolution of the π0 recipe
Where the original π0 model focused on dexterity within a known setting, π0.5 emphasizes transfer to homes and environments the model was never trained in. Physical Intelligence has published some of the more detailed technical reports in the space, making this entry a good reference for the flow-matching approach to action generation.
More in Robot Foundation Models
Skild Brain
A robotics foundation model from Skild AI, founded in 2023 by researchers Deepak Pathak and Abhinav Gupta.
Octo
A transformer-based diffusion policy pre-trained on 800,000 robot episodes from the Open X-Embodiment dataset.
Figure Helix
End-to-end vision-language-action model from Figure that runs entirely on the Figure 02 and Figure 03 humanoids. A dual-system architecture (System 2 ~7B VLM for reasoning, System 1 ~80M transformer for fast control) enables 35-DoF whole-upper-body control. Helix 02 (Jan 2026) extends to whole-body autonomy.
Let's get in touch with us
At the intersection of innovation and technology, we are pioneers crafting a landscape for the digital age.