ChatGPT context window calculator
Estimate how much source context you can give ChatGPT while leaving room for instructions and a useful answer.
GPT-6 Astra budget planner
GPT-6 Astra supports a 1.05M-token context window, up to 922K input tokens and 128K output tokens. OpenAI's most capable model. Astra Pro is the same model served at a higher reasoning effort and shares these limits.
Estimated cost
Source: OpenAI documentation. Last reviewed 2026-09-08. Prices reviewed 2026-09-08.
Frequently asked questions
What models does this calculator cover?
It shows GPT-6 Astra, OpenAI's most capable model, alongside the three GPT-5.6 tiers: Sol, Terra, and Luna. All four have a 1.05M-token context window and a 128K output limit, so the difference between them is price and capability rather than room. Select the variant you are using to plan your context budget.
Why does my cost jump on a large bundle?
OpenAI bills prompts over 272,000 input tokens at 2x the input rate and 1.5x the output rate, and the surcharge applies to the whole request rather than just the tokens above the line. Staying under 272K is often cheaper than trimming elsewhere. Some clients also expose a smaller effective window than the API limit, so check what your tool reports before assuming you have the full 1.05M.
Can I use this for ChatGPT Projects?
Yes. Use it to estimate whether your source text, prompt, and expected answer have enough room before adding files to a project or chat.
Is this an exact OpenAI tokenizer?
No. It is a planning calculator. For exact API tokenization, use OpenAI's model-specific tokenizer tooling.
Can Riflet prepare ChatGPT context files?
Yes. Riflet exports text or Markdown bundles that can be pasted or uploaded into ChatGPT when the product surface supports it.