What the token counter does
LLMs don't read characters or words — they read tokens, and every API bills by them. This counter estimates how many tokens a piece of text uses for GPT, Claude, and Gemini, alongside its character and word counts, and shows a rough cost so you can see what a prompt or document will cost before you send it.
How to count tokens
- Paste your text — a prompt, a system message, or a document.
- Read the token estimate per model family (tokenizers differ slightly between providers).
- Use the count to stay within a model's context window and to estimate cost.
Why token counts matter
Two reasons: cost — you pay per token, so a bloated prompt is a recurring tax — and context limits — every model has a maximum window; exceed it and the request fails or silently truncates. Counting before you send lets you trim, chunk, or pick a larger-context model deliberately. To cut a prompt's token count, try the prompt compressor; to price a workload, the LLM pricing calculator.
