Sources: Mirendil, founded by former Anthropic researchers to build self-improving AI, is in talks to raise ~$1B led by Kleiner Perkins at a $5B valuation
Context & Ripple Effects
Mirendil was first reported in March to be seeking $175 million at a $1 billion valuation for AI models aimed at scientific research, then was reported in June to have closed a $200 million seed round at the same valuation while describing self-improving AI for open-source developers. The newly reported terms would represent a sharp repricing only months after that funding benchmark.
The company’s connection to former Anthropic researchers places the proposed round within a wider capital market that has rewarded frontier-lab pedigrees; Anthropic itself was reported in May to be weighing a much larger financing at a far higher valuation.
First-order effects
- If completed on the reported terms, Mirendil would gain roughly $1 billion to fund its self-improving-AI effort, with Kleiner Perkins taking a leading investor role.
- The proposed $5 billion valuation would set a new private-market benchmark for Mirendil, five times the valuation reported for its June seed round.
Second-order effects
- The proposed repricing gives investors evaluating other early-stage AI-model companies a much higher comparable for teams with frontier-lab backgrounds, raising the premium attached to technical talent and research ambitions.
- Kleiner Perkins’ reported lead role would put a major venture firm at the center of Mirendil’s next financing, rather than leaving the company valued solely on its earlier seed-round terms.
Third-order effects
- If similarly large rounds continue to accrue to a small set of frontier-model startups, access to multibillion-dollar private valuations may become a prerequisite for competing for researchers and long-horizon model development.
- That pattern would concentrate AI-model development among companies able to secure repeated, outsized financings, rather than distributing capital evenly across the broader developer-tool market.
The trend: Frontier-AI investing is increasingly concentrating capital behind young labs whose teams and ambitions resemble established model developers.