LLM context window comparison

A model's context window is the maximum number of tokens — roughly ¾ of a word each — it can consider at once, covering your prompt, any documents, and its own reply. Bigger windows let a model read whole codebases, books or hours of transcript in one pass.

Advertised limits are not the whole story: accuracy often drops toward the end of very long contexts, and some providers cap consumer apps below the API maximum. Figures below are the largest officially documented window per model.