Prime Intellect Raises $130M to Help Enterprises Train Custom AI Agents

Enterprise AI infrastructure gets a major boost

Prime Intellect has closed a $130 million Series A at a $1 billion valuation to expand its platform that enables enterprises to train their own AI agents without depending on frontier labs like OpenAI or Anthropic. Radical Ventures led the round, with backing from Nvidia Ventures, Intel Capital, Dell Technologies Capital, Iconiq, and founders including Perplexity's Aravind Srinivas and Box's Aaron Levie, according to TechCrunch.

Founded in 2024, the startup addresses a growing enterprise concern: how to develop specialized agentic systems while maintaining control over proprietary data and avoiding vendor lock-in. The company provides what it describes as a full-stack solution encompassing compute access, reinforcement learning frameworks, and evaluation tools—essentially packaging the capabilities of a top-tier AI lab into modular, accessible infrastructure.

Why it matters

Enterprises face mounting risks when building on closed AI models. Recent incidents like Anthropic's sudden shutdown of its Fable model have exposed the fragility of third-party dependencies. Companies also worry about sharing sensitive business data with frontier labs that could eventually compete against them. Prime Intellect's approach lets organizations own their AI development stack entirely, a shift that could reshape how enterprise AI is built and deployed.

Reinforcement learning makes custom training viable

The technical foundation enabling Prime Intellect's model centers on reinforcement learning techniques that iteratively reward successful task completion while penalizing errors. This methodology allows companies to refine models for specific business workflows—something that would have been prohibitively complex just a few years ago.

The platform operates as a marketplace where customers select only the components they need rather than committing to an all-or-nothing system. David Katz, a partner at Radical Ventures, noted that while other vendors offer fragments of this capability, Prime Intellect stands out by delivering frontier-level AI lab functionality as a unified offering at accessible price points.

Early traction with enterprise customers

The startup has already attracted customers including Ramp, Zapier, and Flapping Airplanes, who pay for hosted access to its tools. This adoption has driven Prime Intellect to an annualized revenue run rate of $100 million.

Ramp provides a concrete example of the platform's value. The fintech company used Prime Intellect to build an agent for extracting answers from spreadsheets. According to Ramp co-founder and co-CEO Karim Atiyeh, the resulting system outperformed frontier models on accuracy while running faster and at significantly lower cost.

Democratizing AI development

Prime Intellect co-founder and CEO Vincent Weisser frames the company's mission as democratizing access to advanced AI capabilities. He argues that the ability to train sophisticated models shouldn't be concentrated among a handful of San Francisco-based labs, but should extend to every enterprise and nation state.

The details of Prime Intellect's funding and customer traction were first reported by TechCrunch.