The formula
Estimated tokens ≈ characters ÷ 4 (blended with words × 1.33). Input cost = tokens ÷ 1,000,000 × input price. Output cost = tokens ÷ 1,000,000 × output price. Actual tokenisation depends on the model's tokenizer and language.
Questions
How many tokens is a word?
In English, roughly 0.75 words per token, or about 4 characters per token. Code and non-English text typically use more tokens per word.
Is this token count exact?
It is a close estimate using a characters-and-words heuristic. The exact count comes from the specific model's tokenizer, which this lightweight tool approximates without uploading your text.
Is my text uploaded anywhere?
No. All counting and cost estimation happens locally in your browser; nothing is sent to a server.