Figure AI Research

Research

  • Figure AI introduces Helix 2.5 with zero-shot generalization across 30 homes - 2026-09-17 - Research Publication - Holding task-specific data, architecture, training and evaluation fixed, Index pretraining alone raised zero-shot success from 9% to 56%, and a single fixed checkpoint per task was used across all 30 homes with no… - source: figure.ai
  • Figure AI launches Project Go-Big for internet-scale humanoid pretraining - 2025-09-18 - Research Publication - Figure announced Project Go-Big as a large-scale humanoid pretraining initiative built around egocentric human video collected across Brookfield environments. The company says Helix achieved zero-shot human-video-to-robot… - source: figure.ai
  • Figure AI shows Helix folding laundry as a deformable-object manipulation benchmark - 2025-08-12 - Research Publication - On 12 August 2025, Helix folded towels fully autonomously with an end-to-end neural network after the same model had handled logistics package reorientation. The same Helix architecture and training hyperparameters were… - source: figure.ai
  • Figure reports Helix logistics gains on deformable packages - 2025-06-07 - Research Publication - Helix extended logistics handling beyond rigid boxes to deformable polybags and padded envelopes, with temporal visual memory and force feedback helping the robot adapt. In the company's comparison, average handling time… - source: figure.ai
  • Figure AI introduces Helix VLA model for generalist humanoid control - 2025-02-20 - Research Publication - Figure introduced Helix as a generalist vision-language-action (VLA) model that combines visual perception, language understanding, and learned control for humanoid manipulation. The architecture separates a 7–9 Hz… - source: figure.ai

Articles

View Figure AI overview