AI Startup Nous Research Raises $75 Million in Funding

Nous Research, a startup building AI agent software, is in talks to raise at least $75 million in new funding at a $1.5 billion valuation, according to three people with knowledge of the deal reported by TechCrunch. Robot Ventures is leading the round, with support from USV and other investors. The company and investors have declined to comment on the deal.

This funding round is happening remarkably fast. Nous Research closed a previous $50 million funding round less than three months ago at a $1 billion valuation. Jumping from $1 billion to $1.5 billion in roughly 90 days signals how aggressively investors are chasing AI agent companies right now — particularly those with evidence of real user adoption.

What Nous Research Actually Does

Nous Research was founded in 2023 by Jeffrey Quesnelle, Karan Malhotra, Ryan Teknium, and Shivani Mitra. The company has now raised $120 million total from investors including Paradigm, Robot Ventures, North Island Ventures, OSS Capital, and Balaji Srinivasan.

The company's main product is called Hermes. Hermes is an AI agent — think of it as software that can follow instructions and make decisions on its own, similar to how a personal assistant might handle research or routine tasks for you. What matters about Hermes is that it is built as open-source software, meaning anyone can download the code and use it freely.

Hermes has gained real traction. The project has roughly 214,000 stars and nearly 40,000 forks on GitHub, the platform developers use to collaborate on and share code. Those numbers mean thousands of developers are actively using and building on top of Hermes. The software comes with built-in abilities — web search, coding help, image understanding — and is designed to learn new skills automatically as people use it, rather than shipping with a fixed set of features.

Nous Research makes money in two ways. First, it offers a paid cloud-hosted version of Hermes for $20 to $200 per month, letting people use the software without setting it up themselves. Second, it runs a decentralized compute network where users can contribute their own computers to help train and run these AI systems. The decentralized network reflects a crypto-influenced approach to how AI infrastructure could be built and owned.

The Unusual Mix

What makes Nous Research distinctive — and worth understanding — is that it is operating in three different worlds at once: as an open-source project with genuine grassroots adoption on GitHub, as a crypto-native platform raising on token valuations, and now as a traditional venture-backed startup attracting mainstream investors like Robot Ventures and USV. Few companies try all three at once, and each pulls investors with different expectations for how the company should behave and how they might make money.

The real question here is whether the speed of this funding round reflects something real about user demand or simply capital chasing a hot category. A 50 percent valuation increase in three months, without a major new product or clear revenue milestone announced alongside it, is worth thinking carefully about. Nous Research's GitHub numbers show that developers actually use Hermes. Whether that developer interest converts into the kind of paying customers needed to justify a $1.5 billion valuation is the bet this round is making.

It is also important to note that TechCrunch's reporting relies on three unnamed sources, not statements from the company or investors. Both Nous Research and the lead investors have declined to comment or not responded to requests for comment. In the fast-moving AI funding world, valuations discussed mid-negotiation can shift before a deal closes. Readers should treat the $1.5 billion figure as where talks stood at the time of reporting, not as a done deal.

What This Money Will Likely Go Toward

The company has said the new funding will go toward expanding Hermes and its business model, though specifics are unclear. Nous Research has previously published technical work on systems that would allow Hermes to refine itself based on how real people use it — the kind of engineering needed for an AI agent to learn and improve over time without needing constant human intervention. That work likely gives us clues about where the company is headed.

For a market crowded with AI agent startups pitching on demos and test results alone, Nous Research has something more concrete: a GitHub community that genuinely uses its software. Whether that community translates into sustainable revenue, and whether the decentralized compute network actually scales alongside the paid product, are the questions any investor in this round is betting will work out. In a competitive field where many startups look similar on the surface, real user adoption tends to be what separates the winners from the rest.