Paste a prompt or a reply to see its characters, words and an estimate of its tokens, the unit ChatGPT and the OpenAI API use for limits and pricing.
AI models do not read characters or words. They read tokens: pieces of words, whole words or punctuation. Limits on how much you can paste, how long an answer can be and what the API charges are all set in tokens. OpenAI gives a rule of thumb for English: one token is about 4 characters or about three quarters of a word, so 100 tokens is about 75 words.
This page shows both estimates. When they differ, the real count is usually between them. Code, numbers, URLs and non-English text often use more tokens per word. For an exact count use OpenAI's own tokenizer for the model you call. ChatGPT does not publish a fixed character limit for messages: the limit is the model's context window, which differs by model and plan.
About 750 words of English, by OpenAI's rule of thumb of 0.75 words per token.
About 1,333 tokens of English text. Code and other languages often use more.
There is no fixed character limit. ChatGPT limits messages by tokens, and the limit depends on the model and your plan. Long pastes may be cut off or refused.
No. The count runs in your browser and the text is not uploaded.