Mirendil announced a $1 billion Series A round on Nov 12 2026. The round was led by Kleiner Perkins and Clearlake Capital, with investors including Google and several venture firms. The money will fund tools that let one scientist use AI agents to do coding and experiments that normally need a whole team. In a recent test, a researcher trained Mirendil‑Base, a coding model, from scratch in 3.5 months using under $10 million of compute, then added a 10 million‑token context window and launched a private‑beta agent platform called Lor.
Why it matters
The funding could let individual researchers run large AI projects without needing a big lab, speeding up scientific and tech development.