What is a Context Window?

A context window is the amount of text that an AI can "see" and work with at one time. Think of it as the AI's memory during a single conversation or task.

A Simple Explanation

Imagine you're reading a book, but you can only see one page at a time through a window. Everything on that page is clear and usable. But once you turn the page, you can't see what was on the previous page anymore.

A context window works similarly. The AI can read and understand everything within its context window, but once new text comes in and the window "shifts," older information may be lost or pushed out.

How Context Windows Are Measured

Context windows are measured in tokens, not words. A token is a small piece of text—usually a word or part of a word.

For example:

  • ChatGPT has a context window of up to 128,000 tokens
  • Claude has context windows ranging from 100,000 to 200,000 tokens
  • Older AI models might have context windows of just 4,000 tokens

Since tokens are smaller than words, a 100,000 token context window can hold approximately 75,000-80,000 words. That's roughly a short novel or several lengthy documents.

Why Context Matters

The size of the context window determines what the AI can remember and reference in a conversation.

With a larger context window, you can:

  • Have longer conversations without the AI forgetting earlier points
  • Upload longer documents for analysis
  • Ask the AI to review more information before answering
  • Have more complex, multi-part discussions

With a smaller context window, conversations get limited more quickly, and you might need to summarize or restart if you want to reference much earlier discussion.

What Happens When You Hit the Limit?

When you reach your context window limit, different AI systems handle it differently.

Some systems will simply stop accepting more text and tell you that you've reached the limit. Others might "forget" the oldest parts of the conversation to make room for new information.

This is why very long conversations with AI can feel like the AI is losing track of things mentioned at the beginning.

Practical Examples

Example 1: Long Document Analysis

If you upload a 50-page document (roughly 15,000 tokens) to an AI with a 128,000 token context window, you still have about 113,000 tokens left for your questions and the AI's responses. You can ask multiple detailed questions about the document.

Example 2: Detailed Conversation

If you're having a detailed back-and-forth discussion, each message and response consumes tokens. A 20,000-token conversation leaves 80,000 tokens for further discussion in a 100,000-token context window.

How This Changed Over Time

Early AI models like GPT-3 had context windows of just 4,000 tokens. This meant conversations couldn't get very long or complex.

Modern AI models have much larger context windows—some now exceeding 200,000 tokens. This is one of the biggest improvements in recent years, allowing for much more sophisticated interactions.

The Simple Explanation

A context window is the amount of text an AI can "see" and work with at any given moment. Larger context windows allow for longer conversations, bigger documents, and more complex interactions. When you exceed your context window, the AI may forget earlier parts of the conversation. Modern AI models have much larger context windows than earlier generations, making them better for extended discussions and detailed analysis.