Anthros Labs, a startup building evaluation infrastructure for LLM applications, has raised $40 million in Series B funding led by Lightspeed Venture Partners. Existing investors Index Ventures and Conviction participated.
The round brings Anthros's total funding to $63M and values the company at $320M post-money, the team confirmed. Proceeds will fund a 40-person hiring push focused on enterprise customer success and a planned expansion into agent-evaluation primitives.
Anthros's platform lets teams version evaluation suites alongside model versions, run regression tests across model upgrades, and surface drift in production output quality. The product is used by AI teams at several large public companies, with ARR reportedly growing 6x year-over-year.
"Eval is the last mile of every serious AI deployment," wrote one of the founders in the announcement. "Teams used to ship and pray. Now they treat model upgrades like database migrations, versioned, tested, reversible."
The company is hiring across engineering, applied research, and field engineering. See open roles or the product page for more.
