Token Counter

A token counter estimates how many tokens a large language model will read from your text. This free tool gives an approximate count using the splitting rules most BPE tokenizers follow, shows how much of a context window you would use, and estimates cost from the price per million tokens you enter. It is an estimate, not an exact model tokenizer.

Free, no signup. Runs entirely in your browser; files and text are not uploaded.

How to use the Token Counter

  1. Paste your prompt or document into the text box.
  2. Read the estimated token count, characters and words.
  3. Set the context window (for example 128K or 200K) to see the percentage used.
  4. Enter your model’s current price per 1M input tokens to estimate cost.

Features

  • Approximate token count modelled on BPE pre-tokenisation (words, digit groups, punctuation, CJK)
  • Character, word and characters-per-token statistics
  • Context-window usage bar with 8K, 32K, 128K, 200K and 1M presets or a custom size
  • Cost estimate from your own price per 1M input tokens
  • Clear note that results are estimates, with advice on exact counting
  • Runs locally: prompts are not sent anywhere

Frequently asked questions

How accurate is this token counter?

It is an estimate. For ordinary English prose it is usually close to real tokenizer counts, but code, non-English text and emoji can differ more because every model family uses its own tokenizer. For exact numbers use the provider’s tokenizer or token-counting API, such as OpenAI’s tiktoken or Anthropic’s token counting endpoint.

How many characters are in a token?

For English text a token is about four characters, or roughly three quarters of a word, on average. Other languages and code usually need more tokens per character. The tool shows the characters-per-token ratio for your text.

Why doesn’t the tool list model prices?

AI providers change prices and release new models often, so a built-in price list goes out of date quickly. Enter the current input price from your provider’s pricing page to get an up-to-date cost estimate.

Is my prompt sent to an AI provider?

No. The estimate is calculated in your browser with no network requests, so confidential prompts and documents stay on your device.

Related tools

  • Word Counter – Count words, characters, sentences and paragraphs with reading time, readability and keyword density
  • Text Summarizer – Summarize articles, reports and notes with AI in short, medium or long form
  • Keyword Extractor – Extract the main keywords from any text with AI, with counts and density
  • Article Rewriter – Rewrite and paraphrase text in five styles and four tones with AI

Built and reviewed by Naqash Thaheem, Technical Project Manager. Last reviewed 2026-09-24.