Your Yave account will be signed out on this device, along with the other Yave services you use here.

Token & cost calculator

Estimate the tokens in your text and what each model would cost

Your text

Cost by model

Prices as of Anthropic 2026-09 · OpenAI 2026-10 · Google 2026-10 · Prices change often. If the amount matters, check the provider's pricing page.

Model Input Output Total

Estimated tokens

0

This is an estimate, not an exact count. The real number depends on each model's tokenizer — expect it to land inside the range above.

Assumptions

Raise the call count to see the total for sending the same prompt repeatedly.

Add your own model

Add a model that isn't listed, or your contracted rate, and compare it alongside. Stays in this browser only.

About this tool

Estimate how many tokens your text will be and what it will cost on each model. It counts characters and words too, and the text-tidying tools sit on the same screen. What you paste never leaves the browser.

How to use it

  1. 1 Paste your text.
  2. 2 Enter the output tokens you expect and how many runs, and the per-model cost appears as a table.
  3. 3 If you need to, tidy the text in the next tab to bring the token count down.

Good to know

  • Token counts are an estimated range, not an exact figure
  • Prices are as of the last time a person checked them

Frequently asked questions

Why not give me the exact token count?

Models count tokens differently and some do not publish how, so a single number would be wrong. You get an upper and lower bound instead, which is enough to budget with.

The model I use is not in the list.

You can add it yourself. Take the input and output price per million tokens from the provider's pricing page, enter them, and it joins the table.

Does Korean use more tokens than English?

Usually, yes. English packs several characters into one token, while Korean, Chinese and Japanese characters can each take a token or more.

Tools that pair well with this

Esc