Prime Intellect Secures $130M Series A to Let Enterprises Build Their Own AI Agents

Prime Intellect, a startup founded in 2024, has secured $130 million in a Series A funding round that values the company at $1 billion. The investment highlights a growing appetite among enterprises for building their own AI agents rather than depending on a small number of frontier AI laboratories.

The substantial round was led by Radical Ventures, with participation from Nvidia Ventures, Intel Capital, Dell Technologies Capital, and Iconiq. A notable group of angel investors who are themselves founders of prominent companies also joined the round, including Aravind Srinivas of Perplexity, Aaron Levie of Box, Winston Weinberg of Harvey, Jeff Wang of Cognition, and Brendan Foody of Mercor.

Empowering Companies to Become Their Own AI Labs

Prime Intellect's core mission is to give organizations the capabilities they need to train their own agentic AI systems, freeing them from reliance on frontier AI labs. While this ambition would have been difficult to realize just a few years ago, the company points to the rise of reinforcement learning techniques as a key enabler. These methods iteratively reward successful task completion and penalize errors, allowing companies to refine models for specific business tasks and effectively become their own AI lab.

Yet even as it becomes technically possible to bypass closed AI labs, the underlying infrastructure remains so complex that most companies lack the expertise to assemble the necessary components into a production-ready system. That is the gap Prime Intellect aims to fill.

The startup has built what it calls a "full-stack" platform for AI agent development. This includes access to computing power, a reinforcement learning framework, and evaluation tools. The platform functions as a marketplace, offering modular access so customers can pick and choose the specific tools they need without being locked into an all-or-nothing system.

David Katz, a partner at Radical Ventures, said Prime Intellect has stitched these elements together in a way that operates at the frontier while remaining affordable. While other providers offer fragments of the stack, Katz said, Prime Intellect stands out by delivering the capabilities of a top-tier AI lab as a "one-stop shop" for development.

Real-World Adoption and Revenue Growth

The startup's approach has already drawn paying customers, including Ramp, Zapier, and Flapping Airplanes, which pay for a hosted version of its tools. This rapid adoption has propelled Prime Intellect to an annualized revenue run rate of $100 million.

The growth is driven by tangible results. Ramp, a fintech company, used Prime Intellect to build an agent that helps find answers inside spreadsheets. Karim Atiyeh, Ramp's co-founder and co-CEO, said in a statement that the result beat frontier models on accuracy while running at faster speeds and a fraction of the cost.

Growing Wariness of Frontier Model Dependencies

Another significant factor behind Prime Intellect's momentum is a recent shift in how companies view their dependence on frontier AI labs. Businesses are increasingly unwilling to provide proprietary information to companies like OpenAI and Anthropic, citing the risk of losing control over their data. They are also wary of relying on models that could be turned off without warning — a concern underscored when Anthropic's Fable was shut down the previous month.

Katz articulated a deeper worry: companies may be partnering with AI providers that could eventually seek to replace them or generalize into their areas of business. These concerns, he said, are prompting organizations to rethink how they can own their enterprise intelligence and avoid such risks.

Vincent Weisser, Prime Intellect's co-founder and CEO, believes enterprises are actively looking to move away from closed-source frontier models. His company, he said, provides the infrastructure to make that transition feasible. In his view, the capability to train AI models should not be confined to a select few in San Francisco but should be accessible to every enterprise and every nation state.

As enterprises weigh the trade-offs between convenience and control, Prime Intellect's combination of modular tooling, reinforcement learning expertise, and compute access positions it at the center of a broader shift toward self-reliant AI development. The $1 billion valuation signals strong investor confidence in that vision. If you found this article valuable, share it with your network and join the conversation about the future of enterprise AI.

Source: TechCrunch AI