Developer ToolsFree, no sign-up
Context Window Calculator
See how much conversation, document, or code fits in a model's context window.
Pricing dataset updated Aug 28, 2026
200K total window · max output 64K tokens
Usable prompt tokens
196,000
- Approx. words
- 147,000
- Approx. pages
- 294
- Approx. lines of code
- 19,600
- Cost to fill once
- $0.196
Assumes ~0.75 words per token and ~10 tokens per line of code.
How it works
The context window bounds input plus output for a single request. Reserving room for the response matters: a 200K window with a 64K max output leaves 136K for your prompt.
Long-context models remove engineering work but raise per-request cost. Retrieval is still cheaper than stuffing the window on every call.
Worked example
A 200K window
200,000 tokens is roughly 150,000 words, about 300 pages of text, or a mid-size codebase of around 20,000 lines.