gfs.ai

Figure AI/Consumer

Figure 03

The best-funded humanoid, now aimed at your house.

AnnouncedHome pilots · 2026
  • From a $39B-valued company — the field's richest bet
  • Fingertip sensors register about 3 grams of force
  • Reoriented from the factory floor to the home
Non-gendered

Realism index

38/100

Our editorial estimate of how close this machine is to a lifelike human companion — across face, skin, movement and mind. How we score

Figure 03 is what a humanoid looks like when the company building it has almost unlimited money and a point to prove. Backed by a $39 billion valuation, it's the most polished general-purpose machine here — and the clearest signal that the smart money now sees the home, not just the warehouse, as the prize.

The home-focused details are genuinely thoughtful: soft, washable textile covers; foam padding to protect fingers from pinch points; wireless charging through coils in the feet, so the robot simply steps onto a pad to refuel. Underneath, fingertip tactile sensors detect around three grams of force — enough to handle something fragile without crushing it — and the in-house Helix AI lets a fleet of robots pool what they learn.

What it isn't, yet, is buyable, or a companion. Figure 03 has no face and no pretense of one; its warmth is in its engineering, not its expression. Home availability is a late-2026 target for select partners, and there's no public price. It's in this directory as the benchmark for capability — the machine most likely to actually fold your laundry — against which the companions built to be loved can be measured.

Our verdict

The case for

  • The best-funded effort in the field ($39B valuation)
  • Genuinely engineered for homes — soft, washable, safe
  • A proven prior generation on a real BMW line

The case against

  • Not for sale; home availability is a 2026+ target
  • No public price
  • Home autonomy still largely unproven outside demos

What it can do

  • Dexterous fine manipulation with grip- and slip-detection
  • Built for domestic environments
  • Fleet learning — robots share skills through Helix