How many tokens is 100 words?

On average, 100 words works out to roughly 130 to 150 tokens, though the exact number can vary depending on which AI model's tokenizer you're using and the complexity of the text itself. Tokens aren't the same as words, they're smaller units that can represent whole words, parts of words, or even punctuation marks. Common, simple words like 'the' or 'cat' usually count as a single token, while longer or less common words, technical terms, or words with unusual spelling often get split into multiple tokens. Punctuation, spacing, and even capitalization can affect the token count slightly. As a rough rule of thumb that most people in the AI space use, one token is roughly equivalent to about 0.75 words in English, which means 100 words translates to approximately 133 tokens on average. If your text includes a lot of numbers, code, or non-English characters, the token count can run noticeably higher since those often require more tokens per character. If you need an exact count for a specific model, tools like OpenAI's tokenizer let you paste text and see the precise breakdown.