How to use it
- Type the tokens of a typical request (input and output) and how many requests you expect a month.
- Read the list, cheapest first. A model with a dated price change shows what the month would cost after the change.
- Open the Claude and Gemini pages for the tables behind the numbers and the notes on what is not included.
What it does not tell you
It does not say which model does your task well. It has no benchmarks. It leaves out caching prices, long-prompt surcharges (except Gemini 2.5 Pro's lower tier), audio and image prices, and OpenAI's models, whose pages we could not read.
Are these the real prices?
They are what the makers' pages showed on 14 November 2026. Check the pages: prices change, and some have dates written into them.
Why is Claude Haiku 4.5 in the list but not Claude Haiku 5.5?
The page shows a range with a surcharge for long prompts for Haiku 5.5, which the calculator does not model.
This is an independent guide. It is not connected with Anthropic, Google or any company named here. Models and prices change quickly: every figure was read on 14 November 2026 on the maker's own page named under Sources. There are no quality scores here, only what the makers publish. Check the pages again before you decide.