Analysis
Mirendil is in talks to raise new funding at a $5 billion valuation, with Kleiner Perkins in talks to lead and Andreessen Horowitz also in discussions to participate in a round that could bring in up to $1 billion in new capital, according to Bloomberg. Nothing has closed, and the reported terms could still change before signing.
Two Anthropic Alumni, One Quarter
Mirendil was founded in December 2025 by Behnam Neyshabur, who previously co-led Anthropic's scientific AI reasoning team, and Harsh Mehta, a former senior research scientist at Anthropic. The company is building AI meant to accelerate AI research itself, aiming to let open-source developers and scientists develop their own specialized in-house models in fields like medicine and materials science, rather than depending on a general-purpose frontier lab for every use case.
“Bloomberg's own reporting frames this explicitly as talks in progress, not a signed term sheet.”
Three months ago, Kleiner Perkins and Andreessen Horowitz led a $200 million seed round for Mirendil at a $1 billion valuation, with Nvidia also participating, according to Tech Funding News. A five-fold valuation increase in a single quarter is an extreme outlier even by 2026's compressed AI fundraising timelines, where two-to-three-times markups on hot labs within a quarter have become common but a 5x jump remains rare.
A Crowded, Unproven Category
Mirendil sits in a crowded "AI for AI research" category alongside Sakana AI and Reflection AI, and competes indirectly with the internal automated-research efforts every major frontier lab -- OpenAI, Anthropic and Google DeepMind among them -- is already running for itself. Mirendil's pitch is doing this as an independent, model-agnostic vendor rather than a proprietary internal tool, a differentiated position but one that has not yet been tested against paying customers at scale.
The risk in these reported terms is straightforward: the deal has not yet closed, and a five-fold markup on a three-month-old seed round with no disclosed revenue or product metrics is a valuation built almost entirely on team pedigree and investor conviction rather than demonstrated commercial traction. Bloomberg's own reporting frames this explicitly as talks in progress, not a signed term sheet.