Mirendil Inc. has secured $1 billion in a Series A funding round led by Kleiner Perkins and Clearlake Capital. The round included participation from over six other investors, including Google LLC. The startup intends to use the capital to accelerate the development of its artificial intelligence software.
Founded last year by former researchers from Google DeepMind and Anthropic, the company is led by CEO Behnam Neyshabur and CTO Harsh Mehta. Their primary product, Lor, is a platform designed to automate model training tasks by distributing workloads across hundreds of AI agents. The system can automatically detect and fix errors during the evaluation phase.
The company recently demonstrated its technology by having a single researcher build a coding model, Mirendil-Base, in 3.5 months at a cost of $10 million. AI agents performed over 99% of the coding work, including infrastructure setup and data collection. This model serves as the foundation for Fern, a specialized coding tool released last month.
Fern supports a context window of 10 million input tokens, which the company states is ten times the capacity of Claude Opus 5.5. The underlying Mirendil-Base model currently features between 100 billion and 500 billion parameters. Mirendil is also developing upgraded versions of this architecture with parameter counts ranging from 500 billion to 10 trillion.
