Guide
The 200K / 272K context cliff
Cross the context threshold and the premium applies to the entire request — not just the overflow.
Where the cliff is
For Claude and Gemini the threshold is 200K tokens; for OpenAI it is 272K. Above it, input is roughly 2× and output roughly 1.5× — applied to the whole request, retroactively, not only the tokens past the line.
Why it surprises people
A request that creeps over the threshold does not cost 'a little more for the extra tokens' — every token in it is repriced. JP the Cat carries the >200K tiers in its pricing table, so a large-context session is costed correctly.
Common questions
- Is only the overflow billed at the higher rate?
- No — crossing the threshold reprices the entire request at the premium tier.
See your real numbers with JP the Cat
A light, local menu-bar meter for Claude Code, Codex, Grok, and Muse Spark — real cost, honest limits.
Download MeowPrefer Homebrew? Install the same signed, notarized app from JP's public tap.
brew tap 8bittts/jpthecat
brew trust 8bittts/jpthecat
brew install --cask jpthecat