To estimate tokens, paste your text into the Token Counter to get character-based and word-based estimates of how many LLM tokens it will use. These are approximations, since each model has its own tokenizer, but they are good for planning prompt size and cost.
Follow along with the free Token Counter, which runs in your browser, or browse more tools in this category.
How to use the Token Counter
- Paste your prompt or document.
- Read the character-based and word-based token estimates.
- Compare the estimate with your model’s context limit.
- Trim or split the text if you are near the limit.
Worked example
A rule of thumb is roughly 4 characters or three-quarters of an English word per token, so a 1,000-word passage is around 1,300 tokens.
Tips and common mistakes
- Treat the result as an estimate; the model’s own tokenizer gives the exact count.
- Non-English text and code often use more tokens per word.
- Remember that both your input and the model’s output count toward the context window.
Frequently asked questions
Is this the exact token count?
No. It is an estimate because every model tokenizes text differently.
What is a token?
A chunk of text, often part of a word, that a language model reads and generates.
Is my text uploaded?
No. The estimate is calculated in your browser.