Analysis
Vals AI, a San Francisco startup that builds independent benchmarks for testing AI models on real-world professional tasks, raised $40 million in a Series A round led by a16z at a $400 million valuation, according to Tech Funding News. Existing investors 8VC and Bloomberg Beta returned, joined by new investors HRT Ventures and Next Ladder Ventures.
Vals AI's pitch runs counter to how most AI labs market model capability -- rather than academic leaderboards or lab-published benchmark scores, the company evaluates models against tasks pulled from actual professional workflows: finance, law, and other domains where errors carry real financial or compliance consequences. Its research has found that frontier models fail roughly 52% of the real finance-analyst tasks it tests them against, a figure that stands in sharp contrast to the near-saturated scores most frontier labs report on standard academic benchmarks like MMLU or GPQA.
Independent measurement as its own category
The company's growth numbers are the strongest signal in the round: revenue has grown eightfold compared with all of 2025, its customer base has doubled, and its team has tripled over the past six months. That trajectory reflects a structural shift in how enterprises, labs and governments are buying AI -- self-reported benchmark scores from OpenAI, Anthropic, Google and the rest of the frontier field have become less trusted as a basis for procurement decisions, particularly after several labs faced criticism this year for benchmark methodology that inflated real-world capability claims.
Vals AI's bet is that independent, task-specific evaluation becomes required infrastructure for any enterprise deploying AI in a regulated or high-stakes function, the same way independent auditors became required infrastructure for financial reporting once self-reported numbers alone stopped being sufficient for institutional trust. Competing directly in this category are a handful of smaller evaluation startups and academic benchmark consortia, none of which have yet raised at a comparable valuation -- Vals AI's round effectively establishes it as the best-funded independent player in AI evaluation heading into a year where enterprise AI spending scrutiny is intensifying rather than easing.
