AI Charts is a model-comparison and measurement resource that helps users assess how AI systems perform relative to their cost and token use. Its interactive charts plot published benchmark scores against estimated task costs and output tokens, making it easier to compare model configurations and identify which options offer the strongest scores within different budgets. The site covers a broad range of providers and models, with views for model efficiency, coding agents, and capability versus cost.

AI Charts also provides benchmark listings, leaderboards, recent model records, and a release radar that tracks new announcements from first-party sources. Its pages explain the data, benchmark methodology, chart axes, and how results are assembled, and users can download the underlying data as JSON. The site distinguishes its general intelligence-index results from coding-agent measurements and notes limitations in coverage and sourcing. For teams evaluating their own systems, AI Charts offers a local collector to measure agents’ token use. Together, these features support practical comparisons of AI capability, efficiency, and changing model availability.