Vals AI has raised $40 million in a Series A round led by Andreessen Horowitz at a $400 million valuation, with existing investors 8VC, Pear VC and Bloomberg Beta participating and HRT Ventures and Next Ladder Ventures joining as new backers. Founded by Rayan Krishnan and Langston Nashold, the company positions itself as a third-party evaluator and "scorekeeper" for large models, testing how they perform on real business tasks rather than relying on public benchmarks. Vals works with experts across legal, finance, healthcare and coding to score model outputs in practical professional scenarios. It says its private test sets are run only in limited fashion to reduce the risk that model developers train directly to the questions or optimize for leaderboard performance, a problem that can make public benchmark results less useful for enterprise buyers. The funding adds momentum to a growing market for independent AI evaluation as companies seek better ways to judge whether frontier systems can handle production work. Vals says its approach is designed to measure applied usefulness in business settings rather than headline scores on static public tests.