Paradigm leads $50 million Series A round for decentralized AI project Nous Research

Prominent crypto venture capital firm Paradigm has led a $50 million Series A round for the decentralized AI startup Nous Research.

This latest funding round was almost entirely financed by Paradigm, Fortune initially reported . It brings Nous Research's yet-to-launch token valuation to $1 billion, Tommy Shaughnessy, co-founder of Delphi Digital and founding partner at Delphi Ventures — a previous investor in the project — told The Block.

Founded in 2023 by CEO Jeffrey Quesnelle, Head of Behavior Karan Malhotra, Head of Post Training "Teknium" and Shivani Mitra, Nous Research is an open-source AI lab focused on advancing foundational AI models through a decentralized, community-driven approach, offering an alternative to platforms like OpenAI and DeepSeek. It plans to leverage blockchain infrastructure — primarily Solana — to coordinate global contributors, incentivize people to contribute their own idle computing power with tokens and ensure transparency in training processes compared to costly, centralized data center models.

Nous Research previously announced the closure of its initial $5.2 million seed round in January 2024, led by Distributed Global and OSS Capital, with participation from notable investors such as Balaji Srinivasan. An additional $15 million, that includes a previously unannounced June 2024 round, brought its total seed funding to around $20 million, with contributions from Delphi Ventures, North Island Ventures, Together AI CEO Vipul Reddy and Solana co-founder Raj Gokal, according to a statement shared with The Block. Nous Research also received a grant from Andreessen Horowitz at some point, according to a LinkedIn post.

"Nous Research is pioneering truly open AI at scale," Shaughnessy said. "[Its flagship language model] Hermes 3 has surpassed 50 million downloads, powering agents across platforms like X, Telegram and gaming environments. With DisTrO enabling distributed training over the internet and with the coming launch of Psyche on Solana, Nous is redefining decentralized AI infrastructure."

Crypto native team

"We very much came from a mentality that we want to create and serve the world's best AI," Nous Research co-founder Karan Malhotra told Fortune. "This open, community-oriented approach is a powerful contrast to the closed, centralized efforts from incumbent labs," Paradigm Partner Arjun Balaji added. 

Malhotra noted that the AI community has generally been skeptical of crypto, citing its reputation for attracting "grift." However, he described the project's founding team as "crypto native," though they were initially cautious about incorporating blockchain until their research breakthrough in distributed training.

"We don't want to get kind of bogged down by the traditional view of how crypto operates when we're a very serious research lab and an academic lab," he told Fortune. "[But] this is really the only way in which we can make such a massive training run and such a democratic thing possible."

Paradigm, co-founded by Coinbase’s Fred Ehrsam and former Sequoia partner Matt Huang, is one of the largest crypto venture firms, having closed a third $850 million fund last year. This is not Paradigm's first foray into the sector, however, having previously led an $18 million Series A round for decentralized AI training platform Vana.

Psyche Network testnet

Nous Research also announced the testnet for its Psyche Network on Friday — an open infrastructure built on Solana that aims to democratize the development of advanced AI.

The Psyche Network connects diverse GPU hardware to collaboratively train censorship-resistant models in a fault-tolerant way, enabling decentralized machine learning as an alternative to today’s closed and costly cloud-based AI systems.

Nous Research plans to deploy the fresh capital toward expanding its research and development initiatives, building out foundational compute infrastructure for the Psyche Network, strategic hires across its engineering teams and accelerating the development of new AI models.