Generalist Reportedly Adds Nearly $200M at $3B Valuation for Robot Learning From Demonstrations

Generalist has reportedly added nearly $200 million in an 8VC-led Series B extension, reaching a reported $3 billion valuation. Its Gen-1.5 model is built around a different route to robot automation: learning a task from demonstrations instead of requiring custom programming for each job.

A larger round, inside the same Series B

The new capital is described as an extension of Generalist’s $400 million Series B, announced in June at a $2 billion valuation. The extension brings the round’s reported total to $600 million. TechCrunch tied the additional amount to a regulatory filing; SiliconANGLE, citing an Axios source, put the financing at $200 million and said unnamed existing investors joined 8VC.

The deal terms remain reported rather than company-confirmed: Generalist and 8VC did not respond to TechCrunch’s request for comment. The company’s June round included Nvidia and Bezos Expeditions, among other participants.

A brief demo versus a long training run

Generalist is developing a foundation model for various robots. Gen-1.5 targets robotic arms and, the company says, can learn from hand demonstrations captured by robot cameras, sensors attached to a user’s hands, or clips of simulated robots. It also says new tasks can be learned from video demonstrations lasting three to 12 seconds.

What Generalist has disclosed

• Across 10 sample tasks, Generalist reported a 59% average completion rate from one example, rising to 83% with a few additional examples.
• Gen-1.5 took more than eight months to train, according to Generalist.
• The company is working with a handful of customers and using their feedback to tailor the model for specific use cases, according to a person familiar with the company.

The unresolved production test

The $3 billion reported valuation sits alongside early, company-reported evidence for a model intended to reduce custom robot coding. The figures show gains from more examples, but they do not answer how Gen-1.5 performs across customer deployments.

Founded in 2024 by former Google DeepMind researchers Pete Florence and Andy Zeng and former Boston Dynamics engineer Andrew Barry, Generalist is pursuing a foundation-model approach for robots. The reported extension gives that approach more capital; the next material evidence will be whether the demonstration-based method holds up in the company’s customer-specific work.