Model comparison · 2 to 4 models

Compare what the sources actually report.

Every column uses the same Epoch AI catalog record. A LMArena rating appears only for an exact model match in the pinned source snapshot. There is no combined or universal score.

Choose models

Select two to four. The URL updates as you choose.

2 models selected

Swipe or scroll sideways to inspect every selected model.

Shared comparison fields: Epoch AI Data on AI Models, retrieved 2026-07-21, CC-BY, artifact sha256:80ad08198183693101ceb246a5ea272d64c06c147fc41f72cca8526b32483424. The LMArena rating is a separate source observation, not part of a composite. Context and price rows, where shown, are OpenRouter platform listings retrieved 2026-07-22, in USD per million tokens; they are routed platform prices, not first-party provider prices.

Canonical head-to-head pages

Deep-dive comparisons with sourced verdicts.

Each pair page carries a dated verdict, side-by-side sourced facts, and an explicit statement where the sources do not overlap. Pairs exist only where both models share at least one evidence dimension.