Model and Inference / TOKEN WIKI
The capabilities of the
context window
In one sentence
The context window is the range of Tokens that the model can accommodate at one time, usually involving input and output budgets.
Why it matters
nominal window capacity does not equate to reliable utilization of all information at all locations, nor does it represent unlimited memory.
Practical considerations
Evaluates the impact of positioning critical information in long documents and allows space for output, templates, and tool results.
FAQ
Is the context window equal to the permanent memory of the conversation?
No, the application needs to re-provide or retrieve the required information.
