AI Token Estimator

What it is for

For anyone who wants to know, before sending text to an AI, roughly how many tokens it will use and what that costs. You enter a word count and a price per 1 million tokens and get a quick estimate, handy for budgeting summaries, translations or long-document processing.

How it works

The tool starts from a rule of thumb: about 1.3 tokens per word. It multiplies your word count by 1.3 to estimate tokens. Then it divides by 1 million and multiplies by the price per million you entered to get the estimated cost. It never reads the text itself; it only works with the word count you type. Everything runs in the browser, in seconds.

Worked example

A 1,500-word document comes to about 1,950 tokens (1,500 × 1.3). At a price of $10 per million tokens, the estimated cost is 1,950 ÷ 1,000,000 × 10 = $0.0195, which is under two cents. A 150-word email would use roughly 195 tokens.

Limits and caveats

The 1.3 tokens-per-word ratio is an approximation. Real tokenization varies by model and language, and text in Portuguese or Spanish, or with code and symbols, may produce more tokens. Treat it as an order of magnitude.

Frequently asked questions

How many tokens is a word?

On average, close to 1.3 tokens per word in English, per the rule used here. Long or rare words and accented languages tend to use more, which is why the figure is only an estimate.

How do I find the cost of sending text to an AI?

Estimate the tokens, divide by 1 million and multiply by the model's price per million. The tool calculates this from the word count. Remember that the generated reply is billed too.

Is a token the same as a word?

No. A token is a piece of text, which can be a short word, part of a word or a symbol. A word usually becomes one or more tokens, depending on the model and the language.