LIVE — Accepting New Clients 💡 See SEO & Web/App Services →
\n

LLM Token Counter & Cost Calculator

Paste a prompt to count its tokens and see what it costs on each major model - per request and per month.

  • Private – runs in your browser
  • No signup
  • Free, no watermark
Loading tokenizer...
0Tokens (o200k)
0Words
0Characters
–Chars per token
About 375 words per 500 tokens.
Used for the monthly estimate (30 days).
Sorted by cost per request, cheapest first
Model Input tokens Input / 1M Output / 1M Per request Per month

Prices in US dollars per million tokens, standard pay-as-you-go rates, checked against each provider's official pricing page on 4 November 2026. ≈ = estimated token count (the provider does not publish a browser tokenizer). Always confirm on the provider's site before budgeting.

Take it further

AI Chatbots & Assistants Reviews of ChatGPT, Claude, Gemini and more. AI Coding Tools Coding assistants that run on these models. AI Agents Agent platforms - where token costs add up fastest. Chinese AI models DeepSeek, Qwen, Kimi and other low-cost models.

How to use the token counter

  1. Paste your textPaste a prompt, document or sample response. The count updates as you type.
  2. Set expected outputEnter how many tokens you expect the model to reply with, and how many requests you make a day.
  3. Compare costsThe table shows the cost per request and per month for each model, cheapest first.

What is a token?

Language models read text in chunks called tokens - usually a common word, part of a longer word, or a punctuation mark. In English, one token is roughly four characters or three-quarters of a word, so 1,000 tokens is about 750 words. Code, numbers and non-English text usually take more tokens per word.

API providers charge separately for input tokens (what you send) and output tokens (what the model writes back), and output is usually several times more expensive.

How exact are these counts?

The count uses OpenAI's open-source o200k_base tokenizer, the one used by GPT-4o and the GPT-5 family, so for those models it matches the API exactly. Anthropic, Google and most other providers do not publish an in-browser tokenizer for their latest models, and OpenAI has not confirmed the tokenizer for GPT-6, so for those we show a close estimate (marked ≈) - typically within 10-15%. Prices come from each provider's official pricing page; the date they were last checked is shown under the table.

Frequently asked questions

Is my text sent anywhere?
No. Tokens are counted in your browser and no AI model is called, so your prompt stays on your device.
How many words is 1,000 tokens?
About 750 English words. The tool shows words and characters alongside the token count so you can see the ratio for your own text.
Why are output tokens more expensive?
Generating text takes more computing work than reading it, so providers charge more for output - often four to eight times the input price.
Do the prices include cached or batch discounts?
The main table uses standard pay-as-you-go prices. Where a provider offers cheaper cached input, it is shown in the model details. Batch APIs are usually around 50% cheaper again.
How often are prices updated?
We check providers' pricing pages regularly and show the date of the last check. Always confirm on the provider's site before committing to a budget.

Word & Character Counter

Count words, characters and sentences, check readability and platform limits.

Open tool

SERP Preview & Meta Length Checker

See how your title and meta description look in Google, with pixel-width warnings.

Open tool

UTM Link Builder

Build tagged campaign links for Google Analytics, and save your presets.

Open tool

See all free tools →

Updated · Free to use, no account needed.