How the count and cost are worked out
AI models do not read words. They read tokens, small chunks of text that are often a whole short word or part of a longer one. Providers charge per million tokens, with separate prices for the text you send in and the text the model writes back. The calculator uses:
- Input cost = tokens in your text x input price / 1,000,000
- Output cost = expected output tokens x output price / 1,000,000
- Total per request = input cost + output cost
Paste your prompt, set roughly how long you expect the reply to be, and compare the cost of one request across models.
A worked example
A support bot sends a 2,000-token prompt (instructions plus the customer message) and gets back a 500-token reply. On a model priced at $2 per million input tokens and $10 per million output tokens, one request costs $0.004 for input and $0.005 for output, about $0.009 in total. At 10,000 conversations a month, that is roughly $90.
The same traffic on a model priced at $0.10 and $0.50 costs under $5 a month. Many teams send simple requests to a small model and keep the large ones for harder work.
Rules of thumb for English text
- One token is about four characters.
- 100 tokens is about 75 words.
- A typical page of text is 500 to 700 tokens.
- Code, tables, numbers and languages other than English use more tokens per word.
Ways to spend less
Output tokens usually cost four to five times more than input, so asking for shorter answers saves the most. Long instructions that repeat on every request can often be cached by the provider at a much lower rate. It also pays to trim old messages from long chats, since the whole history is sent again with each new turn.
Prices change often. The table above shows the date the prices were last checked. Confirm current rates on each provider's pricing page before you plan a large budget.
