Stats

How it works

1. Type or paste the text you want to analyze.
2. All stats update in real time as you type.
3. Use select or copy to grab the input text for use elsewhere.

Why people use a word counter

Meeting word limits

College essays, cover letters, and social media bios all have word or character limits. Paste your draft here to see the count instantly and trim as needed before submitting.

Estimating reading time

Blog writers and newsletter editors check reading time to keep content at a length their audience will actually finish. The reading time stat here assumes 200 words per minute, which is a standard estimate for adult English reading.

Estimating LLM token costs

If you are working with ChatGPT, Claude, Gemini, or other language models, the estimated token count gives you a rough idea of how much a prompt will cost. One token is roughly 4 characters of English text.

Methodology

Words are counted by splitting on whitespace. Sentences are counted by terminal punctuation (. ! ?) followed by a space or end of text. Text without terminal punctuation counts as one sentence. Paragraphs are blocks separated by one or more blank lines. Reading time assumes 200 words per minute. Token count is estimated at roughly 1 token per 4 characters. This approximation works for most English text across GPT, Claude, and Gemini models. Actual token counts vary by model, language, and vocabulary.

Frequently asked questions

How accurate is the token estimate?

The 1-token-per-4-characters heuristic is approximately 80% accurate for English text. Non-Latin scripts and code may tokenize differently. For exact counts, use your model provider's tokenizer.

Does it count spaces as characters?

The "Characters" stat includes spaces. The "No spaces" stat excludes all whitespace characters.