AI Benchmarks
AI model math rankings
Models are ordered by the reported math index. Use the underlying scores as comparison signals, then open a model record or the methodology for source and limitation details.
0 of 0 models
Failed to load benchmark data. Please try again later.
How to read these AI benchmark rankings
The index values and individual benchmark results shown here are imported from Artificial Analysis. Scores can use different scales, so compare models within the same column rather than treating every number as equivalent. A dash means a score is unavailable.
No benchmark is a universal measure of model quality. Consider your task, cost, latency, context window, and required modalities alongside the reported scores.