Context Window
The context window is the maximum number of tokens a language model can attend to in a single request, covering the system prompt, conversation history, retrieved documents, tool outputs, and the generated response. When inputs exceed this limit, earlier content must be truncated, summarized, or retrieved on demand. Window size shapes how much state an agent can reason over at once.
Definition
The context window is the maximum number of tokens a language model can attend to in a single request, covering the system prompt, conversation history, retrieved documents, tool outputs, and the generated response. When inputs exceed this limit, earlier content must be truncated, summarized, or retrieved on demand. Window size shapes how much state an agent can reason over at once.
Examples
- Fitting an entire 200-page policy PDF into a single 200K-token request.
- Truncating older chat turns once a conversation crosses the model's token limit.
- Using retrieval to inject only the 5 most relevant code files instead of the whole repo.
Related terms
No credit card required
Frequently asked
- What is Context Window?
- The context window is the maximum number of tokens a language model can attend to in a single request, covering the system prompt, conversation history, retrieved documents, tool outputs, and the generated response. When inputs exceed this limit, earlier content must be truncated, summarized, or retrieved on demand. Window size shapes how much state an agent can reason over at once.
- How is Context Window used in practice?
- Context Window is commonly used in scenarios such as: Fitting an entire 200-page policy PDF into a single 200K-token request.; Truncating older chat turns once a conversation crosses the model's token limit.; Using retrieval to inject only the 5 most relevant code files instead of the whole repo..
- How does yno.ai support Context Window?
- yno.ai supports Context Window as part of its agentic AI platform — combining multiple frontier models with extensible tool use.
- What other concepts are related to Context Window?
- Context Window relates to Tokenization, Retrieval-Augmented Generation, Dynamic Context, Large Language Model. Each links to its own glossary page below.