Prime Intellect, a startup that provides computing infrastructure and software tools for building AI agents, raised $130 million in a Series A round announced in July 2026, valuing the company at $1 billion.
The round was led by Radical Ventures, with participation from Nvidia Ventures, Intel Capital, Dell Technologies Capital, and Iconiq. Notable angel investors 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, Prime Intellect offers what it describes as a full-stack platform for AI agent development, covering compute access, a reinforcement learning framework, and evaluation tools. The platform operates as a modular marketplace, allowing customers to select specific components rather than committing to an all-in-one system. Customers including Ramp, Zapier, and Flapping Airplanes pay for a hosted version of the tools, and the company has reached an annualized revenue run rate of $100 million.
The startup’s pitch centers on enabling companies to train their own AI agents without depending on large, closed-source AI labs such as OpenAI or Anthropic. Reinforcement learning techniques — which reward successful task completions and penalize errors — now make it technically feasible for enterprises to refine models for specific business tasks. However, assembling the required infrastructure remains complex enough that most organizations lack the in-house expertise to do it independently.
Concerns about data privacy and model continuity are driving demand, according to David Katz, a partner at Radical Ventures. He noted that companies are wary of sharing proprietary data with frontier labs and of depending on models that could be discontinued — pointing to Anthropic’s recent shutdown of its Fable model as an example.
Ramp co-founder and co-CEO Karim Atiyeh said the company used Prime Intellect to build an agent for finding answers inside spreadsheets, with results that “beat the frontier models on accuracy while running at faster speeds and a fraction of the cost.”
CEO and co-founder Vincent Weisser said the company’s goal is to extend AI training capabilities beyond a small number of large labs. “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.”
Source: TechCrunch