Mirendil has raised $1 billion in a Series A round led by Kleiner Perkins and Clearlake Capital. The San Francisco AI lab builds tools that let software agents carry out frontier research work, and the announcement on November 12 makes it one of the largest Series A financings.
Battery Ventures, Doerr Capital, Gil and Co, Google, M13, Pear VC and S32 joined the round alongside other new and existing investors. Kleiner Perkins co-led Mirendil’s $200 million seed round in June, so the firm has now backed both of the company’s financings.
One researcher, one model, under $10 million
Mirendil’s pitch rests on a result it published on October 21. One researcher, working with agents, pretrained Mirendil-Base, a coding model, from scratch in three and a half months with less than $10 million in compute.
The company then post-trained that model into Fern, which it says holds up to 10 million tokens of context, enough for days of an agent’s work. Its platform Lor, now in private beta, runs hundreds of agents that write code and run experiments for days at a time.
The team that left Anthropic to automate research
Mirendil is led by chief executive Behnam Neyshabur and chief technology officer Harsh Mehta. Neyshabur previously co-led Google DeepMind’s Blueshift team, which worked on reasoning and maths for Gemini, while Mehta built the first version of Anthropic’s internal AI research platform.
They are joined by Shayan Salehian, an early member of xAI, and Tara Rezaei as co-founders. The founding team of 20 researchers and engineers comes from Anthropic, xAI, Google DeepMind and OpenAI.
What the backers are betting on
Andreessen Horowitz co-led the June seed round and its partners summed up the idea in a line that has followed the company since.
Call it vibe research. If it works, it could change the way the AI ecosystem is structured, and support experts across many fields.
Matt Bornstein and Malika Aubakirova, Andreessen Horowitz
The seed round was $200 million at a valuation of about $1 billion, with Nvidia investing afterwards. Mirendil also signed a multi-year partnership with Google Cloud in August to run training and research on its AI Hypercomputer, using Google TPUs and Nvidia hardware.
The company did not disclose a valuation for the Series A. Bloomberg reported in September that Kleiner Perkins was in talks to lead a round of up to $1 billion at a $5 billion valuation, but no source has confirmed that figure as the final terms.
Why the timing matters
The round lands days before Anthropic, where much of the founding team worked, is expected to start trading on Nasdaq. Investors are now paying for the idea that AI can speed up its own development, a capability the largest labs have so far kept in house.
The claims behind the raise are still the company’s own. Mirendil-Base and Fern have no independent evaluation in public yet, and Lor is invitation only. The size of the bet is clear, and the proof will come from the results the company publishes next.
The raise also shows how fast the market for AI research tools is moving. KeeVurds covered ElevenLabs this week as it turned voice AI into enterprise infrastructure, and Mirendil is making a parallel bet that research itself is the next workflow agents will take over.
