Prime Intellect has secured $130 million in Series A funding at a $1 billion valuation to help businesses construct their own AI agents.
Radical Ventures led the round. Nvidia Ventures, Intel Capital, Dell Technologies Capital, and Iconiq also invested. Angel backers include Aravind Srinivas of Perplexity, Aaron Levie of Box, Winston Weinberg of Harvey, Jeff Wang of Cognition, and Brendan Foody of Mercor.
Founded in 2024, the company aims to let organisations train agentic systems without depending on frontier AI labs. Reinforcement learning techniques now allow firms to refine models for specific business tasks by rewarding successful actions and penalising errors. This capability lets companies act as their own AI lab.
Despite this progress, the underlying infrastructure remains too complex for most firms to assemble into a production-ready system.
Prime Intellect fills that gap with a full stack for AI agent development. The offering combines compute access, a reinforcement learning framework, and evaluation tools.
The platform operates like a marketplace. Customers select specific modular tools rather than committing to an all-or-nothing package.
“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,” said David Katz, a partner at Radical Ventures. He noted that while competitors offer fragments, Prime Intellect provides the capabilities of a top-tier AI lab as a one-stop shop.
Customers like Ramp, Zapier, and Flapping Airplanes pay for a hosted version of the tools. This uptake has pushed the company to an annualised revenue run rate of $100 million.
Tangible results drive the growth. Ramp built an agent to find answers inside spreadsheets using Prime Intellect. The result beat frontier models on accuracy while running faster and at a fraction of the cost. Karim Atiyeh, co-founder and co-CEO of Ramp, made that claim in a statement.
Why companies are moving away from closed labs
Another factor is the risk companies face when building on top of frontier labs. Firms increasingly refuse to provide proprietary information to OpenAI and Anthropic due to data control concerns. They are also wary of depending on models that can be suddenly turned off, as happened with Anthropic’s Fable last month.
“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 said. “All of these things are causing people to think, ‘How do I own my own enterprise intelligence and not have these risks’.”
Vincent Weisser, co-founder and CEO of Prime Intellect, believes enterprises want to move away from closed-source frontier models. His company provides the infrastructure for that transition.
“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.”
What it means
For people building tools, the shift is practical. You no longer need to wait for a major provider to release a new feature or worry about your data leaving your control. You can build, test, and deploy systems that match your specific needs without paying premium prices for generic access. The barrier to entry has lowered, but the requirement for technical skill remains.
