All calculators

AI Token Counter

Estimate how many tokens your text uses — helpful for prompt budgeting with LLMs.

0 chars0 words0 sentences0 paragraphs

Start typing to estimate your token count.

Formula

A useful estimate for English text: tokens ≈ words × 1.3, or roughly characters ÷ 4

How to calculate

  1. 1Count the words in your text.
  2. 2Multiply by 1.3 for a quick token estimate in English.
  3. 3Or divide the character count by 4 for a rough equivalent.
  4. 4Remember code and non-English text tokenize less predictably.

How to use this calculator

  1. 1Paste or type your prompt or text.
  2. 2Choose your target AI model family if the estimate differs.
  3. 3See the token count, word count and estimated cost context live.
  4. 4Adjust the text and watch counts update as you type.

Examples

Short prompt

A 100-word prompt is roughly 130 tokens — comfortably inside any modern model's context window.

Long document

A 5,000-word article is roughly 6,500 tokens — check your model's input limit before pasting it in.

Common questions

What exactly is a token?

A token is a chunk of text — often a word or part of a word. 'Calculator' may split into 'cal' + 'culator', which is why tokens usually outnumber words.

Why do estimates vary between models?

Each model uses its own tokenizer with a different vocabulary, so the same text can tokenize differently across providers.

How do tokens relate to cost?

Most AI providers price per token — input tokens and output tokens are billed separately, so both counts matter.

Explore by category