What Figure AI does
Figure AI (Figure) is an AI robotics company developing general-purpose humanoid robots and an embodied AI “vision-language-action” (VLA) stack intended to let humanoids operate in changing real-world environments. The company’s product direction centers on a sequence of humanoid generations (F.01, F.02, and F.03), plus Helix, its proprietary VLA model that Figure describes as controlling the robot’s behavior directly from sensory inputs (and, in Figure’s description, from pixels) to produce long-horizon autonomy. Figure also operates BotQ, its high-volume manufacturing facility, which it credits as enabling scale-up from prototype throughput toward a manufacturing “fleet” approach that increases the amount of real-world data available for continual improvement of Helix.
Business model-wise, Figure appears to pursue commercialization through deployments of humanoid robots into customer environments (e.g., logistics/manufacturing and, separately, home-use pilots implied by the product roadmap). Its recent partner announcements emphasize scaling deployments through enterprise distribution/logistics networks and industrial manufacturing floors, while its most recent technical and roadmap posts emphasize autonomy improvements (for example, full-body control, coordination among robots, and longer-horizon household tasks).
Strategic position (based on its own disclosures) is defined by three linked efforts: (1) growing robot throughput via BotQ and increasing field deployments, (2) scaling Helix training and capabilities using larger real-world datasets, including a partnership-driven “internet-scale” pretraining effort, and (3) raising large capital commitments to fund compute infrastructure, data collection, and manufacturing scale for humanoids.