— Chatbot & AI

LLM Token Counter

Paste any text and get a live estimate of how many tokens it will use — plus word and character counts — right in your browser.

0Estimated tokens
0Words
0Characters
This is an estimate. As a rule of thumb, one token is roughly 4 characters or about 0.75 words in English (~100 tokens per 75 words). Real tokenization varies by model and tokenizer, so exact counts will differ.
— What it does

The LLM Token Counter is a free tool that estimates how many tokens a piece of text will consume in ChatGPT, Claude, Gemini, or other large language models. It blends two common rules of thumb — roughly four characters per token and about 0.75 words per token — into a live estimate, alongside exact word and character counts. Because real tokenization differs by model, treat the result as an estimate, not an exact bill.

  • Live estimate of tokens, words, and characters as you type.
  • Blended heuristic based on the common 4-characters and 0.75-words-per-token rules.
  • Clearly labelled as an estimate, since exact tokens vary by model and tokenizer.
  • Runs entirely in your browser — nothing is uploaded and no text leaves your device.

How to use it

1

Paste your text

Drop a prompt, document, or chat transcript into the input box.

2

Read the estimate

The estimated token count updates instantly, alongside exact word and character totals.

3

Plan around it

Use the estimate to stay within a model context window or to budget API token usage.

Frequently asked

How accurate is the token estimate?
It is an approximation. Real tokenizers split text differently per model — code, punctuation, and non-English text all change the ratio — so treat the number as a close estimate rather than an exact count.
How many tokens is a word or a character?
As a rule of thumb in English, one token is roughly four characters or about 0.75 words, which means around 100 tokens per 75 words. This tool blends both ratios for its estimate.
Why do tokens matter for ChatGPT and Claude?
Models have a fixed context window measured in tokens and most APIs bill per token, so estimating tokens helps you avoid truncated prompts and control cost before you send a request.
Is my text uploaded anywhere?
No. The counter runs entirely in your browser with JavaScript. Your text is never sent to a server or saved, so it is safe to paste sensitive prompts.

Related tools

LLM API Cost Calculator

Estimate what an LLM API actually costs — pick a model, set tokens per call and monthly volume, and see cost per call, day, month and year.

Open tool →

AI System Prompt Generator

Turn a few fields into a clean, structured system prompt — role, context, guidelines, guardrails, and a fallback your assistant can actually follow.

Open tool →

Word & Character Counter

Paste or type text and see live word, character, sentence and paragraph counts — plus reading and speaking time — instantly.

Open tool →

← All free tools

— Built by saavos

These tools are free. So is the first version of your agent.

saavos is the AI agent that lives on your website — themed to match your design, answering visitors from your own content, and telling you what they actually want to know. Paste your URL and see it answer, before you install anything.

Make my site feel alive