Context Window.

The context window is the maximum number of tokens an LLM can consider at once, encompassing both the input prompt and the generated output. Early models had small context windows (4k tokens). Modern models support 128k to 1 million tokens, enabling processing of entire books, codebases, or extensive research documents in a single interaction.

Why Context Window Size Matters

A larger context window allows the model to consider more background information, maintain longer conversations without 'forgetting' earlier content, process longer documents for summarisation or analysis, and receive more examples in few-shot prompts. For content marketing, it means entire long-form articles can be analysed and revised in one pass.

Practical Limits

Even with large context windows, model performance can degrade on information positioned in the middle of very long inputs, a phenomenon called the 'lost in the middle' problem. Critical instructions and key information are best placed at the beginning or end of long prompts rather than buried in the middle.