A concrete example
If you hand over several reports, the model only keeps what fits inside its context window.
Why it matters
It shapes the quality of a long analysis, an agent, and a RAG system.
You'll see it when a tool refuses a PDF that's too long, cuts off a conversation, or advertises 128k, 200k, or 1M tokens.
Don't mix it up with
Token: A token is a small chunk of text the AI counts to measure what it reads, writes, and bills.
RAG: RAG lets an AI answer using documents pulled up at the moment you ask the question.
Common mistakes
- Thinking a large window replaces a good selection of documents.
- Pasting too much text with no hierarchy.
- Confusing long-term memory with the immediate context.
Quick checklist
- First I check whether the word names a concept, a tool, a risk, or a metric.
- I tie it to a concrete case: If you hand over several reports, the model only keeps what fits inside its context window.
- I keep the main trap in mind: Thinking a large window replaces a good selection of documents.
Quick questions
What is Context window in AI?
The context window is how much text a model can take into account in a single conversation or request.
Where will I run into Context window?
You'll see it when a tool refuses a PDF that's too long, cuts off a conversation, or advertises 128k, 200k, or 1M tokens.
Which word should I read next?
Start with Token, RAG, Context engineering.