Enigma raises $71M seed to build foundation models for general-purpose robot control
The AMW Read
New entrant joining a well-known player map (GR00T, V-JEPA) with a data-efficiency claim that could meaningfully update the baseline for robotics foundation-model deployment costs; seed size signals capital-cycle dynamics.
Enigma raises $71M seed to build foundation models for general-purpose robot control
Enigma Ltd., a San Francisco-based startup founded by cybersecurity researchers Jonathan Jacobi and Gal Niv, launched today with a $71 million seed round led by Index Ventures and Ribbit Capital, with participation from Conviction Partners and angel investors from Google DeepMind, Anthropic, and OpenAI. The company is developing foundation models designed to run on any robot, dramatically reducing the amount of training footage needed to adapt an AI to a new machine or environment.
Why it matters: Enigma enters the crowded physical AI foundation-model space at a time when the capital-intensity of training robotics models is still a structural barrier to adoption. Its claim — that it can adapt models to new robots with a fraction of the usual training data — directly addresses the data-collection bottleneck that has kept most industrial robotics AI in the lab. The investor roster (Index, Ribbit, Conviction) and angel-list from DeepMind/Anthropic/OpenAI signal that the frontier lab diaspora is betting on a model-centric approach to robotics, not just on hardware or teleoperation stacks.
Our take: Enigma is the latest entrant in a segment that already includes Nvidia’s GR00T (which also uses natural-language config) and Meta’s V-JEPA, but the company’s claim of “any robot” compatibility is an aggressive positioning that, if validated, could challenge the hardware-specific fine-tuning moat that incumbent robotics vendors rely on. The seed-round size is notable for a company with no prior revenue or product track record — it reflects the capital compression pattern in which investors are willing to write large early checks to secure a pole position in what they believe will be the next foundation-model frontier. The real signal will be whether Enigma can translate its seed capital into a production-grade model that outperforms existing open-weight robotics baselines across diverse hardware form factors.
