Prime Intellect just raised $130 million at a $1 billion valuation, betting that corporations shouldn't have to hand over their proprietary data to a handful of Silicon Valley labs to build artificial intelligence systems. The startup, founded two years ago, is selling a different vision: tools that let companies train their own AI agents in-house, without depending on closed-source models from OpenAI, Anthropic, or other frontier labs.
The funding round was led by Radical Ventures, with participation from Nvidia Ventures, Intel Capital, Dell Technologies Capital, Iconiq, and angel investors including Aravind Srinivas of Perplexity, Aaron Levie of Box, Winston Weinberg of Harvey, Jeff Wang of Cognition, and Brendan Foody of Mercor. The backing reflects a growing concern among enterprise leaders: relying on proprietary AI systems means surrendering control over sensitive business information and remaining vulnerable to sudden service disruptions.
The Problem With Depending on Giants
Right now, most companies lack the technical infrastructure to build sophisticated AI agents on their own. The underlying systems are complex. They require specialized expertise in reinforcement learning—the technique that iteratively rewards successful task completion and penishes errors—plus access to significant computing power and evaluation tools. Assembling all of this into a production-ready system remains out of reach for the vast majority of enterprises.
That dependency creates real risks. David Katz, a partner at Radical Ventures, pointed to concrete concerns driving companies away from reliance on frontier models. "Companies increasingly don't want to provide proprietary information to OpenAI and Anthropic because of the risk of losing control over their data," Katz said. They're also wary of depending on models that can be suddenly turned off—Anthropic's Fable service was discontinued last month, leaving customers scrambling. "How do I know that I'm not working with a company that is going to try to replace me and generalize to what I'm doing," Katz asked. "All of these things are causing people to think, 'How do I own my own enterprise intelligence and not have these risks'."
What Prime Intellect Is Building
Prime Intellect's platform functions as a marketplace. It provides compute access, a reinforcement learning framework, and evaluation tools—but customers can pick and choose what they need instead of being locked into an all-or-nothing system. The company is essentially democratizing access to the infrastructure that, until now, only the largest AI labs could afford to build and maintain.
Vincent Weisser, Prime Intellect's co-founder and CEO, framed the mission in explicitly political terms. "It shouldn't just be a few nerds in a glass tower in San Francisco that have the capability to train AI models," he told TechCrunch. "It should be every enterprise, every nation state."
The platform is already attracting customers. Ramp, Zapier, and Flapping Airplanes pay for hosted versions of Prime Intellect's tools. The adoption has pushed the company to an annualized revenue run rate of $100 million. Ramp used the platform to build an agent that searches through spreadsheets for financial answers. Karim Atiyeh, Ramp's co-founder and co-CEO, said the result "beat the frontier models on accuracy while running at faster speeds and a fraction of the cost."
David Katz emphasized that Prime Intellect's value proposition isn't just technical—it's about providing what used to be exclusive capabilities at scale. "They've stitched this together and built it in such a way that they're operating at the frontier in a way that's affordable," he said. "While others offer bits and pieces, Prime Intellect is unique in providing the capabilities of a top-tier AI lab as a 'one-stop shop' for development."
Why This Matters:
This funding round reflects a fundamental question about power in the AI era: Should advanced AI capabilities remain concentrated in a few private companies, or should enterprises have the tools and infrastructure to build their own systems? The risks of concentration are real. Companies that depend entirely on proprietary models lose control over their data, face sudden service disruptions, and remain vulnerable to being replaced by the very platforms they rely on. Prime Intellect's approach—making enterprise-grade AI development infrastructure accessible and modular—addresses these concerns by shifting power away from centralized gatekeepers. If the company succeeds in its mission, it could reshape how organizations deploy AI, reduce their dependence on closed-source systems, and create more competition in a market currently dominated by a small number of players. That shift matters not just for individual companies, but for broader questions about who controls the technology that increasingly shapes business, governance, and society.