Series A Funding and Strategic Investors
Mirendil, a San Francisco-based artificial intelligence firm, has secured $1 billion in a Series A funding round, according to a company announcement on Wednesday. The investment was led by Kleiner Perkins and Clearlake Capital, with participation from several other venture capital firms and strategic investors, including Battery Ventures, Doerr Capital, Gil & Co, M13, Pear VC, and S32. Notably, Google, the parent company of Alphabet Inc., also participated in the round.
The capital injection is intended to fund the company’s research and development efforts aimed at democratizing frontier AI R&D. Mirendil’s stated objective is to provide scientists and engineers with the tools necessary to conduct high-level research, potentially reducing the institutional scale required to develop advanced models.
Technical Developments and Agentic Workflows
The funding follows recent technical milestones reported by the company regarding the efficiency of agent-driven development. Mirendil reported that the use of AI agents allows smaller teams to achieve results previously reserved for larger organizations.
As part of these recent developments, the company highlighted the pretraining of “Mirendil-Base,” a competitive coding model. According to company data, a single researcher utilized agents to pretrain the model from scratch in 3.5 months, utilizing less than $10 million in compute resources. This model was subsequently post-trained into “Fern,” a model capable of handling up to 10 million tokens of context, which the company notes is sufficient to encompass several days of an agent’s continuous work.
Product Roadmap and “Lor” Beta
Mirendil is currently developing “Lor,” a platform currently in private beta. The company describes Lor as a system that enables hundreds of agents to write code and execute experiments autonomously for extended periods. Mirendil stated that its own internal operations rely on the Lor platform for daily tasks.
The company intends to use the new capital to scale these capabilities, moving toward a model where individual researchers or small teams can perform tasks that historically required large institutional resources. Mirendil remains a small team focused on making frontier AI R&D more accessible to the broader scientific and engineering communities.
