Prompt Token Estimator
Get a rough token and context-size estimate for a prompt without transmitting its content.
Runs in your browser. No signup or cloud computer.
61 / 100,000
Result
Your result will appear here. The input stays in this browser tab.
How to use it
Paste a prompt to estimate its approximate token count using a simple text heuristic.
Try this example
Explain how Matrix keeps agents working when a laptop closes.
What to check
Actual tokenizer counts vary by model and format; use provider metering for billing.
Review the result before using it in a published page, application, or agent configuration.
Related free tools
Need this work to keep going?
Matrix gives agents a private cloud computer for files, connected tools, and work that continues after you close your laptop. The browser tool above is free; a Matrix computer requires a paid plan.