HUMANOID.INTEL
FIG.AIFigure 03+1.5B.SeriesCTSLA.BOTOptimus+8k.Units.Q1BOS.DYNAtlas+New.CEO.2026AGIL.ROBDigit+AMZN.ScaleNEURA4NE-1+$1B.SeriesDAPP.TRONApollo+$520M.ExtAUNIT.REEG1+3k.Ships.25SUNDAYHomeBot+$1.15B.ValGALBOTG1+RMB2.5B.SerBSANC.AIPhoenix+$90M.SeriesD1X.TECHNEO+$125M.SerCMIND.ROBStealth+Founded.2026FUND.YTD2026$5.8B.Raised
Terminal/Brain DB/Helix
productionVLAProprietary

Helix

by Figure AI
Architecture
VLA
Parameters
Undisclosed
Training Data
BMW factory teleoperation data
License
Proprietary
Status
PRODUCTION
Open Source
No
Robots Supported
Figure 02Figure 03
About

Figure AI's proprietary vision-language-action model powering autonomous humanoid behavior in factory settings. Helix integrates multimodal perception with real-time motor control, enabling Figure robots to perform complex manipulation tasks with minimal human intervention.

Key Differentiator
End-to-end VLA trained on real-world BMW manufacturing data at production scale
Funding Context

$2.6B total funding at $6.7B valuation (Series B, Feb 2025)

Milestones
2024-03Initial Helix model demonstrated on Figure 01
2024-08Figure 02 deployed at BMW Spartanburg with Helix
2025-02Series B at $6.7B valuation
2025-07Figure 03 launch with Helix v2
2025-11BMW scales to 100+ unit deployment