Figure AI (Figure)
Developing general-purpose humanoid robots capable of autonomous labor and interaction with the physical world
Integrating large models with whole-body motion control; first mover in commercial deployment (BMW partnership) and rapid iteration of software and hardware capabilities
Company Overview
Figure AI (Figure) was founded in 2022 and is headquartered in Sunnyvale, California, USA. The company is currently privately held. It focuses on developing general-purpose humanoid robots designed to replace humans in repetitive, dangerous, or labor-intensive tasks through AI-driven whole-body motion control and autonomous decision-making. Within the AI industry chain, Figure occupies the "Embodied Intelligence" segment under the "Terminal & Interaction Entry (C-end)" layer, making it one of the core carriers for bringing large models into the physical world. Since its founding, the company has completed multiple funding rounds, raising over $750 million from top technology companies and investors including OpenAI, Microsoft, NVIDIA, and Jeff Bezos, with a valuation that once reached $2.6 billion. In 2024, Figure entered into a commercial partnership with automaker BMW, deploying Figure 01 robots at a BMW factory for material handling and component assembly trials—marking the first large-scale commercial deployment of humanoid robots.
Figure currently operates a single business line: the R&D, manufacturing, and sales of general-purpose humanoid robots. Its robot products have the following characteristics:
- Fully autonomous motion capability: Uses AI models (including self-developed end-to-end neural networks and integrated OpenAI vision-language models) to perceive the environment, plan paths, grasp objects, and complete complex actions in real time.
- Hardware specifications: The Figure 02 is approximately 1.7 meters tall, weighs about 60 kg, has 41 degrees of freedom (articulated joints), a 9.1 kg payload, and can work continuously for about 5 hours (2.25 kWh battery).
- Application scenarios: Starts with manufacturing plants as the base deployment scenario (material handling, assembly assistance), with future expansion into logistics sorting, warehouse management, retail services, and home services.
Figure has not yet disclosed official volume pricing; it is expected to adopt a robot-as-a-service (RaaS) model or a one-time purchase plus maintenance contract model. The company's revenue model relies primarily on robot sales/leasing income and subsequent software upgrade services.
Technology Moat
Moat 1: AI-Driven Whole-Body Motion Control and Autonomous Decision-Making
Figure deeply integrates large language models and vision-language models (VLMs) into whole-body robot control. The robot captures environmental and voice commands via a monocular camera and microphone array, with neural networks directly generating joint control commands—eliminating the traditional "perception-planning-control" pipeline. This end-to-end approach greatly improves operational fluidity and robustness. For example, Figure 01 can autonomously identify a target, plan a grasping posture, and complete the handover in response to conversational commands such as "hand me that red box." In addition, the Figure team has built from scratch low-power, high-performance motors, reducers, and sensors, achieving coordinated optimization of hardware and algorithms.
Moat 2: First to Achieve Commercial Validation
Unlike competitors (e.g., Boston Dynamics, Tesla Optimus) that are still in the lab or pilot-production stage, Figure signed a partnership with BMW in 2024 and is conducting long-term real-world training in a factory environment. This first-mover advantage allows Figure to accumulate real industrial-scenario data, accelerate model iteration, and build customer trust. The BMW partnership also provides Figure with key revenue validation and a template for scaled deployment.
Moat 3: Top-Tier Investors and Ecosystem Resources
Figure's investors include OpenAI (providing large model capabilities), NVIDIA (providing compute chips and simulation platforms), and Microsoft (cloud computing), forming a complete ecosystem of "AI models + compute + cloud." This alliance enables Figure to access cutting-edge AI capabilities quickly and lower R&D costs.
| Dimension | Data |
|---|---|
| Global humanoid robot market (2030E) | ~$38 billion (Goldman Sachs forecast) |
| Industry competition stage | Early stage: multiple companies developing in parallel; no oligopoly yet |
| Key competitors | Boston Dynamics (Spot/Atlas, commercial quadruped focus), Tesla (Optimus, mass-production potential), Unitree Robotics (H1, motion performance focus), 1X Technologies (incubated in Norway) |
| Downstream customers | BMW (signed), other automotive/manufacturing companies (in discussion), logistics and e-commerce companies |
| Figure funding rounds | Latest Series B ($675 million raised in 2024, valuation of $2.6 billion) |
Financials & Growth
| Metric | Data |
|---|---|
| Revenue (2024 estimate) | ~$5–10 million (primarily from BMW partnership trial revenue) |
| Gross margin | Not disclosed (expected negative early, turning positive with mass production) |
| Net margin | Significant losses (mainly R&D expenditure) |
| Cumulative funding | >$750 million |
| Core growth logic | 1) Scale BMW program to hundred-unit deployment in 2025–2026; 2) Acquire a second automotive OEM customer; 3) Launch Figure 03 and reduce unit cost below $50,000; 4) Expand into logistics and services |
Key Risks:
- Technology maturity risk: The long-term reliability and safety of humanoid robots in complex, unstructured environments have not been fully validated, and critical failures could erode customer trust.
- Commercialization pace risk: The BMW partnership is still at the trial stage, and large-scale mass production has not begun; any delay or cancellation of follow-on orders would significantly hurt revenue expectations.
- Intensifying competition risk: Tesla plans to mass-produce thousands of Optimus units in 2025, Boston Dynamics has shifted Atlas to electric power, and Chinese companies such as Unitree and Zhiyuan are accelerating their catch-up efforts. Figure faces pressure from potential price wars or technology gap closure.
- Capital burn risk: Humanoid robot R&D consumes cash at a very high rate (hundreds of millions of dollars per year). If follow-on funding is not secured in time or commercialization underperforms, Figure could face a cash-flow crisis.
- Regulatory and ethical risk: Deploying robots into factories to work alongside humans raises issues related to safety regulations and labor displacement, potentially leading to policy restrictions and public opinion pressure.
Core Investment Thesis / Industry Value Summary:
Figure is currently the only company in the world to have achieved commercial deployment of humanoid robots. With end-to-end AI whole-body control, BMW endorsement, and a top-tier ecosystem alliance, it has taken an early lead in the golden track of embodied intelligence. If it can complete hundred-unit deployment within 2–3 years and prove economic viability, it is well positioned to become a core beneficiary of the wave of physical AI. However, uncertainties arising from technology iteration and intensifying competition should not be underestimated.