An AI token is a chunk of text—roughly three-quarters of a word—that a language model like ChatGPT processes, and you pay for them with every question you ask and every answer you get.
Think of tokens as the fuel gauge in your car. You don't need to know the chemistry of gasoline to drive, but you do need to know when you're running low. ChatGPT reads your prompt, breaks it into tokens, and then generates a response token by token. The more you write and the longer the reply, the more tokens you burn. This matters because most AI tools have hidden limits.
ChatGPT's free plan cuts you off after a certain number of tokens per conversation. Paid plans give you more, but you're still metered. A 500-word article might cost you 700 tokens.
A rambling 3,000-word story could eat up 4,000 tokens fast. I've seen people get frustrated when the AI just stops mid-sentence. That's usually a token limit hitting a wall.
One practical tip: if you're working on something long, break it into chunks. Don't ask for a 10-page report in one go. You'll hit the ceiling and lose your flow.
Also, different models have different limits. GPT-4 can handle way more tokens than older versions, which means it can remember more of your conversation before it starts forgetting the beginning. For a deeper dive, see our guide on how to write effective prompts that don't waste tokens.