How many tokens is your text?
A 1,000-word English text is about 1,300 tokens with GPT-4o.
Paste yours below and get the exact number, using the real tiktoken encoder — not a characters-divided-by-four guess.
What the numbers mean
- Token
- The unit a language model reads. Roughly 3/4 of an English word, but it varies: code, URLs and non-Latin scripts use more tokens per character.
- BPE (byte pair encoding)
- The algorithm that splits text into tokens. Each model family ships its own vocabulary, so the same sentence gives different counts on different models.
- o200k_base
- The encoding used by GPT-4o and the o-series. Larger vocabulary than the older cl100k_base, so it usually needs fewer tokens for the same text.
- cl100k_base
- The encoding used by GPT-4 and GPT-3.5. Still the right count for older models and for many local models that copy the same vocabulary.
- Context window
- The maximum number of tokens a model can hold in one conversation, counting both your input and its output.
Questions
How many tokens is 1,000 words?
About 1,300 tokens in English with GPT-4o. The exact number depends on the words: technical terms, code and numbers tokenise worse than plain prose.
Is this count an estimate?
No. It runs the real tiktoken BPE encoder for the model family you pick, the same vocabulary the model uses. It is not a characters-divided-by-four approximation.
Why does the same text give different numbers on different models?
Because the vocabularies differ. o200k_base (GPT-4o) is larger and usually more efficient than cl100k_base (GPT-4).
Do you store my text?
No. The text goes to a stateless endpoint, is counted, and is not written anywhere.
Need this on 10,000 documents?
The same counter is a free HTTP endpoint. No key, no signup, 0.6 ms per call.
curl -X POST https://rust.mazzagrp.com/v1/tokens \
-H 'Content-Type: application/json' \
-d '{"text":"your text here","model":"gpt-4o"}'