Figure 03 represents a complete hardware and software redesign aimed at a general-purpose humanoid that can learn and adapt in real-world environments. Announced in October 2025 as Figure's third-generation humanoid, it can autonomously handle household chores (like laundry, cleaning, loading dishwashers) and assist with manual labor in commercial settings. The entire platform is purpose-built to enable Helix, Figure's Vision-Language-Action AI model, enabling it to understand natural language commands and visually assess its surroundings to carry out tasks.
The robot integrates multiple breakthrough technologies: a fully redesigned sensory suite with 8 total cameras (6 main cameras plus embedded palm cameras in each hand), custom tactile sensors in the fingertips that can detect forces as small as 3 grams (about the weight of a paperclip), and an upgraded audio system with repositioned microphone and a speaker that is 2× larger and ~4× louder than Figure 02. The vision system delivers 2× frame rate, ¼ the latency, and ~60% wider field of view per camera compared to the previous model.
Figure 03's design emphasizes safety and home use: it is covered in soft, washable textiles and padded with multi-density foam to eliminate pinch points and reduce the risk of injury or damage. It is notably lighter and more compact than its predecessor (≈9% lower mass and smaller volume than Figure 02) for easier navigation in tight household spaces.
Crucially, Figure 03 is built for autonomy and continuous operation. It supports wireless inductive charging via coils in its feet: the robot can simply step onto a charging pad and recharge itself at 2 kW without plugs. High-speed data offload (10 Gbps mmWave) allows Figure 03 to upload large volumes of sensor data for fleet learning and AI improvement between shifts. These innovations, combined with Helix AI, aim to let fleets of Figure 03 robots learn from each other and operate nearly 24/7 in homes or businesses.