Vals AI is an independent benchmarking platform that puts the latest AI models to the test on real, economically meaningful tasks. Rather than relying on self-reported numbers from AI labs, the team runs its own evaluations across areas like legal research, tax and finance, software engineering, medical coding, and higher-risk domains such as cybersecurity and recursive self-improvement. The results are then published as leaderboards you can browse by model, benchmark, or head-to-head comparison.
