VLA models
Helix
United States
First VLA to output continuous high-rate control of the entire humanoid upper body (wrists, individual fingers, torso, head) and to run fully onboard embedded low-power GPUs, including zero-shot two-robot collaboration with a single set of weights.
OrganizationFigure AI
CountryUS
Release2025-02
Params (B)7
Open weightsno
Architecturedual-system VLA (System 1/System 2): 7B internet-pretrained VLM at 7-9 Hz + 80M reactive visuomotor policy at 200 Hz
Control (Hz)200
Modalitiesvision, language, action
Embodimentshumanoid
Training data~500 h of teleoperated humanoid manipulation demonstrations with auto-generated natural-language annotations
Sources : Figure AI (2025-02-20)Humanoids Daily (2026-07-09)
← Back to comparator · VLA models