Skip to content

AI models

Glossary

What each number in the comparison means, in plain language, and when it matters for your choice.

Benchmarks

GPQA DiamondGPQA Diamond, on the Epoch AI leaderboard (CC BY 4.0, latest version we checked), is 198 biology, physics and chemistry GPQA Diamond (OpenRouter)This is the same GPQA Diamond test, run by OpenRouter Benchmarks (CC BY 4.0, latest version we checked) on the provider OTIS Mock AIMEOTIS Mock AIME, on the Epoch AI leaderboard (CC BY 4.0, latest version we checked), is a set of math problems in the sty FrontierMath (tiers 1 to 3)FrontierMath (tiers 1 to 3), on the Epoch AI leaderboard (CC BY 4.0, latest version we checked), is a set of unpublished SimpleQA VerifiedSimpleQA Verified, on the Epoch AI leaderboard (CC BY 4.0, latest version we checked), asks short factual questions with SWE-bench VerifiedSWE-bench Verified, on the Epoch AI leaderboard (CC BY 4.0, latest version we checked), is 500 real issues from Python r τ-bench Verified (airline)τ-bench Verified (airline), on the OpenRouter Benchmarks leaderboard (CC BY 4.0, latest version we checked), simulates a

Price, context and usage

Price per tokenList price per 1 million input and output tokens on OpenRouter, in US dollars. A token is a piece of text, usually small Context windowHow much text the model can consider at once, adding up the prompt, the documents and the answer, in tokens, according t Open weightsOpen weights mean you can download the model and run it on your own infrastructure. Closed weights mean access is throug OpenRouter usageShare of tokens processed on OpenRouter, according to its public ranking, for the latest day published by that source.Hugging Face downloadsDownloads of the official Hugging Face repository, in that source's 30-day window.