Prime Intellect, a San Francisco‑based startup that sells compute power and specialized software for building AI agents, closed a $130 million Series A financing round on Tuesday. The deal, which lifts the company’s valuation to $1 billion, was spearheaded by Radical Ventures. Nvidia Ventures, Intel Capital, Dell Technologies Capital and a line‑up of angel investors—including the founders of Perplexity, Box, Harvey, Cognition and Mercor—also wrote checks.
Founded in 2024, Prime Intellect’s mission is simple: give large organizations the tools they need to create their own “agentic” systems without leaning on the opaque models of OpenAI, Anthropic or other frontier labs. The firm argues that recent advances in reinforcement learning—where models receive rewards for completing tasks and penalties for errors—make it possible for businesses to run a private AI lab tuned to their specific workflows.
The company’s platform bundles three core components: on‑demand compute, a reinforcement‑learning framework and a suite of evaluation tools. Customers can pick and choose the pieces they need, turning the service into a marketplace rather than 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.
Prime Intellect’s approach has already attracted a handful of high‑profile users. Fintech challenger Ramp used the platform to build an agent that extracts answers from spreadsheets, a task that frontier models struggled with. "The result beat the frontier models on accuracy while running at faster speeds and a fraction of the cost," Ramp co‑founder and co‑CEO Karim Atiyeh told TechCrunch. Zapier and Flapping Airplanes have also signed up for hosted versions of the tools, paying the startup for a managed solution.
Revenue growth reflects that traction. The company now reports an annualized run rate of $100 million, a figure it attributes to the concrete cost savings and performance gains its customers experience. The funding will be used to scale the compute infrastructure, add new reinforcement‑learning capabilities and broaden the sales team to reach more enterprises that are wary of handing proprietary data to third‑party AI providers.
Enterprise leaders are increasingly uncomfortable with the risks of relying on closed‑source models. Recent incidents—such as Anthropic’s abrupt shutdown of its Fable model—have underscored the fragility of a dependency on external labs. Companies also fear that sharing sensitive data with providers like OpenAI could expose them to intellectual‑property loss or regulatory scrutiny. "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, highlighting the shift toward self‑owned intelligence.
Co‑founder and CEO Vincent Weisser believes the market is at a tipping point. "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 fresh capital positions Prime Intellect to become the go‑to infrastructure provider for that vision, offering a one‑stop shop that rivals the capabilities of the biggest AI labs while keeping costs and control firmly in the hands of its customers.
Dieser Artikel wurde mit Unterstützung von KI verfasst.
News Factory APP - agentische News für besseres SEO & AEO.