Free tools · Brain

Token counter: how much of the context your text takes up

Paste a text, a prompt or your assistant's instructions file. The count runs in your browser with the o200k tokenizer and shows the tokens and how much of a context window they take up. Nothing leaves your machine.

Free tool · Brain

Use it now, no account needed

A prompt, a document, or your assistant's instructions file. The count runs here, in your browser.

Tokens

—

o200k tokenizer

Characters

0

Words

0

The number is exact for the o200k tokenizer, used by GPT-4o and GPT-5. Claude, Gemini, and other models split text their own way, so their count differs, but in the same range.

Questions

Common questions

Step-by-step guide on the blog (in Portuguese): Token counter: how much of the context window your prompt takes up (in Portuguese)

Building something with agents?

In the Construindo com IA community, open and free, your question has a place and the answer gets marked as the solution.

Other free tools

More tools in the line

Connectors

Test an MCP server

Paste the address and see if it responds, with which tools, and whether it asks for login.

Open the tool

Assistant

Agent prompt generator

Five answers become an agent's instructions, ready to paste.

Open the tool

Universal MCP

JSON formatter: format and validate JSON

Paste JSON: the error with line and column, or the formatted JSON with fields and types.

Open the tool

Connectors

CPF and CNPJ generator and validator

Generate valid CPF and CNPJ numbers for testing, including the alphanumeric CNPJ, or check a number's digits.

Open the tool

An instructions file is sent in full with every conversation. With Brain, the assistant fetches only what it needs.

Decisions, preferences and where the work left off are kept in a space of your own, and the assistant searches by meaning for what is relevant to the question. You see and fix everything in a dashboard.