What exactly is a 'token' in AI and why does it matter for cost?

A token is a chunk of text that an AI model reads or generates, and it can be as short as one character or as long as one word, with the general rule being that 1 token equals about ¾ of an English word, which matters for cost because most AI tools charge based on the number of tokens you use. Think of it like paying for electricity. You don't pay for "having lights" — you pay for the kilowatt-hours you consume. With AI, tokens are your kilowatt-hours. When you send a prompt, the AI breaks it into tokens to understand it. When it writes a response, it generates new tokens one by one. So a 100-word prompt and a 500-word article will cost you for roughly 750 tokens total. Here's a concrete example: the sentence "I love AI tools." is about 4 tokens ("I", " love", " AI", " tools"). But a more complex word like "uncharacteristically" might be 4 or 5 tokens all by itself. This is why using clear, concise prompts can directly save you money — you're not wasting tokens on fluff. A practical tip: most platforms have a tokenizer tool on their pricing page where you can paste text and see the exact token count. Use it. It's eye-opening to see how punctuation and spaces add up. This also explains why different pricing tiers exist; a plan with a 200,000-token limit per month can handle a lot of blog posts, but if you're processing entire novels, you'll blow through it fast. Understanding tokens turns an abstract monthly bill into a predictable, manageable expense. **Related**: How much does it cost to run an AI writing tool each month? | What's the difference between a token and a word in ChatGPT?