Figure AI

Figure AI is a robotics company building general-purpose humanoid robots powered by AI foundation models. Founded in 2022 by Brett Adcock, Figure reached a $39 billion valuation in its 2026 funding round — making it one of the most valuable private robotics companies in history — with backing from Microsoft, NVIDIA, OpenAI, Intel, and Jeff Bezos.

Helix: Dual-System Architecture

Figure's Helix system represents a novel approach to robot intelligence. Rather than using a single end-to-end model, Helix pairs two specialized networks: a vision-language model (VLM) that understands the scene and a vision-language-action (VLA) model that generates motor commands. The VLM processes camera input and language instructions to build a semantic understanding of the task; the VLA translates that understanding into precise joint-level actions at 200Hz. This dual-system design lets Figure update its reasoning model (VLM) independently from its motor control model (VLA), avoiding the trade-off between intelligence and reaction speed that plagues single-model approaches.

Helix was trained using a combination of imitation learning from human demonstrations, sim-to-real transfer, and reinforcement learning. Figure operates its own teleoperation fleet for data collection, with human operators remote-controlling Figure 02 robots to generate high-quality manipulation data at scale.

Figure 02: Commercial Platform

The Figure 02 is a full-size humanoid robot designed for real-world commercial deployment. It features articulated hands with dexterous manipulation capability, a vision system with multiple cameras for depth and scene understanding, and onboard compute sufficient to run Helix inference locally. Figure has deployment agreements with BMW for automotive manufacturing and is targeting warehouse logistics as its initial beachhead market.

The AI-Native Approach

Figure's strategy differs fundamentally from legacy robotics companies like Boston Dynamics. Where Boston Dynamics built decades of mechanical and locomotion expertise before adding AI, Figure started from the AI side: treating the robot as a foundation model inference platform that happens to have a body. The bet is that AI capability is advancing faster than mechanical engineering, and that good-enough hardware plus frontier AI will beat excellent hardware plus mediocre AI. The $39B valuation reflects investor conviction in this thesis — and in the broader argument that whoever solves general-purpose humanoid intelligence captures one of the largest addressable markets in history.

Figure AI represents the embodied AI frontier of the agentic economy — the extension of autonomous agents from the digital world into physical space, where they can manipulate objects, navigate environments, and perform labor in the real world.