Robotics startup Generalist hits $3 billion valuation just two months after its last raise

Generalist AI, a two-year-old startup building foundation models for robots, has raised roughly $200 million in fresh capital led by the venture firm 8VC, pushing its valuation to $3 billion just two months after its previous round closed at $2 billion, according to people familiar with the deal cited by TechCrunch and separately reported by SiliconANGLE. A notice of exempt securities offering filed with the Securities and Exchange Commission on August 24 independently corroborates the scale of the raise, though neither Generalist nor 8VC had responded to requests for comment as of the TechCrunch report.

The filing shows the San Francisco-based company, incorporated in Delaware under the name Artificial General Dexterity, Inc. before adopting the Generalist AI brand, registered an offering of just over $208 million and had already sold about $198 million of it to 32 investors, with a first sale date of August 7. That extends a Series B that opened in June at $400 million, bringing the combined round to roughly $600 million raised in under three months — an unusually fast sequence even by the standards of the current AI funding market.

The numbers

Generalist's earlier Form D from June, also on EDGAR, shows that round registered just under $400 million, of which roughly $364 million had been sold at the time. Combined with this month's tranche, the company has now drawn in close to $562 million in disclosed sales across the two filings alone, before accounting for whatever remains open in the current offering. The jump from a $2 billion valuation in June to $3 billion now — a 50 percent increase in roughly eight weeks — places Generalist among the fastest-appreciating startups in the crowded field of "physical AI," the industry's shorthand for AI systems designed to operate machines and robots rather than generate text or images.

• New round: roughly $200 million led by 8VC, per sources; SEC filing shows about $198.2 million sold of a $208.2 million offering as of August 24.
• Combined Series B, June to August: roughly $600 million raised at a valuation that rose from $2 billion to $3 billion.
• 32 investors participated in the August tranche alone, according to the regulatory filing.
• Company founded in 2024 by Peter Florence and Andy Zeng, formerly of Google DeepMind, and Andrew Barry, formerly of Boston Dynamics.

How Generalist got here

Generalist was founded in 2024 to build what its founders call "physical AGI" — foundation models that let robots of many different shapes learn new physical tasks the way large language models learn new writing tasks, through pattern generalization rather than task-specific programming. In a post on the company's own blog accompanying the June raise, Generalist said its GEN-1.5 model lets robots pick up new skills from video demonstrations lasting as little as three to twelve seconds, rather than the weeks of task-specific training that has typically been required. "Millions of robots are operating in the world today. Billions more are coming," the company wrote, framing the wager as one on data, models, hardware and infrastructure compounding together rather than any single breakthrough.

The founding team's pedigree — Florence and Zeng both spent years on robotics research at Google DeepMind, while Barry worked at Boston Dynamics, one of the most established names in advanced robotics — has been central to the company's pitch to investors. 8VC, which led the new tranche, first backed Generalist in the June round and laid out its thinking at the time in its own investment writeup.

"If foundation models reshape physical work the way they reshaped software, Generalist will be one of the defining companies of that transition," 8VC wrote in June, citing dexterity — precise, human-like manipulation — as robotics' central unsolved problem and the clearest near-term path to commercial value.

Who is affected

The raise lands amid a broader surge of capital into companies trying to give robots general-purpose intelligence rather than narrow, single-task programming, a wager that investors are increasingly describing as robotics' equivalent of the "ChatGPT moment" for language models. Generalist's investor list spans both AI-native venture firms and strategic technology players: alongside 8VC and Radical Ventures, which led the June tranche, the company has drawn capital from Nvidia, Union Square Ventures, Bezos Expeditions, Spark Capital, Boldstart Ventures and Norwest, along with individual backers including Stanford AI researcher Fei-Fei Li and investor Naval Ravikant.

For robot manufacturers and industrial customers, the more immediate consequence is competitive: Generalist is one of several well-capitalized labs, alongside rivals building similar general-purpose robot models, racing to be the default intelligence layer that hardware makers license rather than build in-house. TechCrunch's sourcing indicates Generalist has kept a low public profile while working with what it described as "a handful" of commercial customers, suggesting the company's near-term ambitions center on proving out real deployments rather than a broad public launch.

What happens next

Generalist has not disclosed how it intends to deploy the new capital, and the company's silence in response to press inquiries leaves open questions about which customers are testing its models and on what hardware. The SEC filing indicates roughly $10 million of the registered offering remained unsold as of the August 24 filing date, suggesting the round could still grow modestly before it formally closes. With GEN-1.5 already released and the company's valuation up 50 percent since June, the next test for Generalist will be converting research momentum and investor enthusiasm into robots that operate reliably outside the lab — the gap that has stalled previous waves of humanoid and general-purpose robotics ventures before commercial deployment could scale.