Figure’s Helix 2.5 Shows the Next Test for Portability

Figure’s Helix 2.5 highlights a new commercialization test for humanoids: whether autonomy can move across bodies and environments without retraining.

Updated

What is this trend?

Helix 2.5 shows that humanoid autonomy can improve sharply in new homes without fresh training, making portability across bodies and sites the next commercialization hurdle.

  • Zero-shot success in unseen homes jumped from 9% to 56% after Index pretraining.
  • The key question is now redeploying one autonomy stack across new robots and environments.
  • Safety, certification, and integration still slow real-world humanoid rollout.
  • Winning vendors will pair embodied data engines with fast adaptation workflows.
  • Operators want supervised pilots that avoid long robot-specific tuning cycles.

What’s the latest?

Figure’s Helix 2.5 adds a commercialization-grade signal: in 30 unseen homes, zero-shot success rose from 9% to 56% after Index pretraining, with no data collection, fine-tuning, or adaptation in those homes.

How it developed

  1. Control Stack Moat, Simulation-to-Edge Deployment, and Lifecycle Services Monetize Robotics
    • Robot Value Is Moving Into the Control Stack
  2. Orchestration Becomes the Value Layer, Throughput Wins Over Labor Replacement
    • AI-native autonomy enabled by edge compute upgrades

Go deeper

Curated long-form picks on this trend — podcasts, videos, and analysis, by vantage.

Stay ahead in Robotics

Get the weekly brief in your inbox — the developments, what they mean by vantage, and what to do next.