Why Token Count exists
Token Count is an independent utility for inspecting AI text quantities and planning token costs. It aims to make the assumptions visible: which engine counted the text, where an estimate is rough, and which price snapshot a calculation used.
Project and contact
The project is maintained through the GitHub account rujieliang-code. You can review the public project or use its available GitHub contact and issue options to report a problem:
For a useful report, include the page, browser, selected model, expected behavior and observed behavior. Include only a short, non-sensitive example when needed. Do not share API keys, confidential prompts or private source files.
Independence and accuracy
This site is not operated by the model providers named in the interface. Product names identify the models and references being discussed. The provider's current documentation, account reporting and invoice remain the references for service capabilities and billing.
Sources and checked dates appear on the new reference pages. The methodology explains counting scope and estimate limitations, and privacy explains the data handling in this website.
This version
Token Count includes a local token counter, a Claude API cost calculator, context budget tools and practical guides. The website is maintained at www.token-count.com.