Tokens to Words Converter
Convert between tokens and an approximate word or character count.
Estimate only, based on ~0.75 words and ~4 characters per token. Real tokenization varies by model and text.
Recommended next steps
Related tools
Estimate text tokens and rough AI API costs before running a prompt.
Estimate the cost of generating images across model presets.
Encode or decode text to and from base64, with a URL-safe variant.
Frequently asked questions
A common rule of thumb for English is that one token is roughly 0.75 words, or about 4 characters. So 1,000 tokens is about 750 words. This is an estimate — real tokenizers split text into sub-word pieces, so the exact count depends on the model and the wording.
Divide the word count by 0.75 (the same as multiplying by about 1.33). For example, 750 words is roughly 1,000 tokens. Switch this tool to "Words to tokens" to do it automatically, including an approximate character count.
Language models use sub-word tokenizers (like BPE), so common words may be a single token while rare words, code, numbers, emoji, or non-English text can be several tokens each. The ~0.75 words-per-token ratio is a useful planning average, not an exact count for a specific model.
No. The conversion and the paste-to-estimate count both run entirely in your browser. Nothing you type or paste is uploaded to a server.
Last updated 2026-06-23.