AI Token Counter & Cost Estimator
Paste any text to estimate its token count and what it would cost to send to popular AI models.
| Characters | 0 |
| Words | 0 |
| Estimated tokens | 0 |
| Likely range | 0 – 0 |
| Model | Input $/MTok | Output $/MTok | Est. cost |
|---|
Cost assumes the model's reply is about as long as your input. Prices are editable defaults — always check the provider's pricing page. Token counts are estimates (±10–15%); exact counts require each model's tokenizer.
How it works
Tokens are the unit AI models actually read and bill by — roughly word fragments, not words. API pricing is quoted per million tokens (MTok) for input and output separately, so knowing your prompt's token count is the difference between guessing your API bill and planning it. The cost column assumes a reply about as long as your input; long-form generation will cost more on the output side. Prices in the table are editable, so you can update them as providers change their rates.
FAQ
How accurate is the estimate?
It uses the standard heuristic that one token is roughly 4 characters or 0.75 words of English text, which lands within about 10–15% of real tokenizer counts for typical prose. Code, non-English text, and unusual formatting can drift further. For exact counts, use the provider's own tokenizer or token-counting API.
Why do models count tokens differently?
Each model family uses its own tokenizer — a learned vocabulary that splits text into pieces. The same sentence can be 100 tokens under one tokenizer and 130 under another, which is why this tool shows a range rather than one exact number.
Is my text uploaded anywhere?
No. Counting and cost math run entirely in your browser with JavaScript. Nothing is sent to any server.