AI Token & Cost Estimator
Paste a prompt or document to estimate how many tokens it uses, then plug in your model's prices to see what each request and a whole month of traffic would cost. Your text never leaves your browser.
Characters
116
Words
17
Estimated tokens
29
Likely range
25–35
Cost calculator
Enter your model's current prices from your provider's pricing page. The defaults are example values only.
Cost per request
$0.00459
Cost per month
$45.87
How the estimate works
For English text, one token is roughly four characters or three quarters of a word. The estimator starts from that rule and adds extra tokens for code symbols and non-Latin characters, which usually split into more tokens. It shows a likely range because every model family uses its own tokenizer, so the exact count varies between providers.
For billing-accurate numbers, use your provider's token-counting endpoint or the usage figures returned with every API response.
Why input and output are priced separately
AI APIs charge for everything you send (system prompt, conversation history, documents) and everything the model writes back. Output tokens are usually several times more expensive than input tokens, so asking for concise answers is one of the easiest ways to cut costs.
Ways to reduce token costs
- Trim the system prompt — it is sent with every request.
- Send only the relevant parts of documents instead of whole files (see RAG explained).
- Summarize or trim long chat histories, which are re-sent on every turn.
- Use prompt caching and batch processing when your provider offers them.
- Use a smaller model for simple tasks like classification and extraction.
Learn more in our guide: What Are Tokens in AI Models?