🧮 AI Token & Cost Calculator

What this tool does

It estimates how many tokens a piece of text becomes when sent to a language model, and what that costs as model input. The estimate is the average of two rules of thumb - characters divided by four, and words times 1.33 - because real tokenizers land between them for ordinary English. The text never leaves your browser.

When to use it

Before pasting a long document into an API call, to see whether it fits a context window. When budgeting a batch job - a thousand articles at 1,200 tokens each is a number you want before running it, not after. When trimming a prompt, to see how much a cut actually saves.

How to read the result

Treat the token count as an order of magnitude, not an invoice. Real tokenizers differ per model, and code, URLs and non-English text tokenize heavier than prose - a text full of JSON can come out 30-50% above this estimate. The cost column covers input only; output tokens are billed separately and usually at a higher rate, so a long answer can cost more than the question.

Limits

The prices in the table are typed in by hand and age from the day they were written - model names and rates change often, so check the provider's own pricing page before relying on a figure. The tool has no per-model tokenizer and no way to know your output length. It answers one question well: roughly what does it cost to send this text in.