Text Token Estimate

Free Text Token Estimate calculator with clear step-by-step results.

About Text Token Estimate

Estimates token count from the actual text you paste in, blending a characters-per-token ratio with a words-per-token heuristic. The ratio varies sharply by content: English prose averages about four characters per token, code closer to three, and CJK scripts under two.

Formulas

Character-based estimate

Tokens = characters / characters per token for the content type

Word-based estimate

Tokens = words x 1.33, blended with the character estimate for non-CJK text

Worked example

With the default values below, Text Token Estimate returns Estimated tokens of 17. Change any input above and every figure here is recalculated in your browser.

Inputs used

Text to estimate: Large language models tokenize text into subword units before processing it.
Language characteristicsone of English prose, Code, European language with accents, CJK (Chinese, Japanese, Korean): English prose

Results

Estimated tokens: 17 tokens
Characters: 76 chars
Words: 11 words
Characters per token assumed: 4.00

Step by step

1Values used
2Character-based estimate
3Word-based estimate
4Estimated tokens= 17 tokens
5Characters= 76 chars
6Words= 11 words
7Characters per token assumed= 4.00

Assumptions and limits

The result is derived only from the 2 inputs listed above. Anything not among them is not part of the calculation.
In particular, this calculator does not take inflation, tax, fees or charges or compounding frequency into account.
Figures are rounded for display to 2 decimal places, so totals built from several rounded parts can differ slightly from the unrounded arithmetic.

How these calculators are built and tested is described on the Methodology page. If a figure looks wrong, please report it.

Frequently Asked Questions

Why is code more token-dense?

Punctuation, indentation, camel-case identifiers and operators fragment into many short tokens, so the same character count yields more tokens than prose.

How accurate is this?

Usually within 10-15% for prose. Only the model's own tokenizer gives an exact count, and different model families tokenize the same text differently.