Figure AI vs Unitree
A side-by-side comparison of Figure AI and Unitree, two Robotics tools, drawn from Ignaite's continuously-verified listings.
Compared from listings verified as of
Figure AI
RoboticsHumanoid robots powered by Helix, an in-house vision-language-action model.
View Figure AIAt a glance
| Attribute | Figure AI | Unitree |
|---|---|---|
| Category | Robotics | Robotics |
| Pricing | PAID | PAID |
| License | Proprietary | Proprietary |
| Deployment | Local | Local |
| Platforms | API | API |
| Model support | Self-contained (on-device) | Self-contained (on-device) |
| Vendor (differs) | Figure AI | Unitree Robotics |
| Capabilities (differs) |
|
|
The honest brief
Figure AI
Exited its OpenAI deal to own the full stack — Helix runs onboard, so hardware and AI are built end-to-end in-house.
- Proprietary Helix VLA model
- Full-stack hardware + AI
- Real pilots (BMW manufacturing)
- Figure 03 targets the home
- Not a buyable product yet
- Capital-intensive, long timelines
- General-purpose autonomy unproven at scale
Unitree
Its robots are buyable today at consumer prices — the R1 humanoid starts at $5,900 — while most humanoid rivals remain pilot-only.
- World's top humanoid seller in 2025
- R1 humanoid from $5,900
- Profitable — rare in robotics
- Quadrupeds, arms, and LiDAR too
- R1 lacks dexterous hands
- About an hour of battery on R1
- Aimed at researchers, not households
When to pick which
Both cover Vision-language-action, Dexterous manipulation, and Bipedal locomotion.
Pick Figure AI if you need Robot foundation model.
- Robot foundation model (primary capability)
Pick Unitree if you need World model.
- World model (secondary capability)
They share capabilities, but each leads with different ones as a headline job:
Figure AI is built around Vision-language-action.
- Vision-language-action (primary capability)
Unitree is built around Dexterous manipulation and Bipedal locomotion.
- Dexterous manipulation (primary capability)
- Bipedal locomotion (primary capability)