LLM Context Window Comparison: Models That Read the Most at Once

It is the amount of text a model can handle in one conversation: everything you send plus everything it writes. Once the limit is reached, older text has to be dropped or shortened.

Roughly 750,000 words, which is several long books. The exact number depends on the language and the text.

No. It means more text fits. Quality on very long inputs differs between models, and long inputs cost more per request, so use only as much as the task needs.

A model appears only when an open source has published results for it. New models can take days or weeks to be measured, and models with no published score are left out rather than guessed.

Prices and context windows come from the open-source LiteLLM price list (MIT licence), which records each model maker's own price page. Prices change often, so confirm on the maker's site before you buy. Epoch scores come from Epoch AI (CC BY 4.0).

We check our open data sources every day and show the date of the data at the top of the page. A model only moves when its source publishes new results, so a number can stay the same for weeks.