Helix
Figure's in-house vision-language-action model for humanoid robot control.
About
Helix is Figure AI's in-house vision-language-action model, unveiled in February 2025, that turns camera input and natural-language instructions into continuous robot motor commands. It uses a two-system design: a 7B VLM planner running at 7-9 Hz and an 80M fast visuomotor controller at 200 Hz, giving high-speed continuous control of a humanoid's full upper body — 35 degrees of freedom — and coordinating two robots on shared tasks from the same weights.
Helix is proprietary and runs on the robot's embedded GPUs; no weights, API, or public benchmarks exist, and the "first" claims (fastest VLA, dual-robot control) are Figure's own. A successor, Helix 02, added a learned whole-body locomotion layer in 2025.
Specs
Context window
Pricing
Free quota
API access
Weights
Released
Benchmarks
Provider-reported unless a source says otherwise. Benchmarks are marketing-adjacent by nature — treat them as directional.
Use cases
Trust score — 6/10
Scored 6/10 (verified tier, metrics-v1). Evidence: Proprietary, no public weights/API/benchmarks; architecture figures are Figure-reported.; Helix 02 successor adds whole-body locomotion.. License: Proprietary; internal to Figure AI robots..
How the trust score is built
Full methodology →- Adoption · stars/downloads, log-scaled vs peers
- 4.0/10
- Maintenance · recency, releases, issue load
- 6.0/10
- Openness · license + docs
- 0.0/10
- Manual review · sanity read + red-flag check
- 8.0/10
- · Proprietary, no public weights/API/benchmarks; architecture figures are Figure-reported.
- · Helix 02 successor adds whole-body locomotion.