Mirendil has raised $1 billion in Series A financing to expand tools intended to make advanced AI research and development accessible to scientists and engineers.
Kleiner Perkins and Clearlake Capital led the round. Participants included Battery Ventures, Doerr Capital, Gil & Co, Google, M13, Pear VC and S32, alongside other new and existing investors.
The company is developing an approach in which AI agents perform coding and experimental work over extended periods, allowing smaller research teams to undertake projects that would otherwise require substantially more personnel.
Mirendil highlighted three recent developments: its Mirendil-Base coding model, the subsequently post-trained Fern model, and Lor, an agent-based research environment now in private beta.
According to the company, one researcher working with agents pretrained Mirendil-Base from scratch in three and a half months using less than $10 million in computing resources.
Fern supports up to 10 million tokens of context, which Mirendil describes as sufficient to hold days of an agent’s work.
Lor extends the approach through hundreds of agents capable of writing code and running experiments for days at a time. Mirendil says it uses Lor in its own daily operations.
Together, the products reflect a strategy of using agents not only as assistants for individual tasks but also as infrastructure for sustained technical work.
The company plans to use the financing to advance that model and broaden access to frontier AI development. Its announcement emphasized enabling individuals and small teams to pursue work traditionally associated with larger organizations.
Mirendil remains a small team and is recruiting people to support the effort. The new financing provides substantial backing as it develops research tools, evaluates its early models and introduces Lor to additional users through private beta.
