antist.ai

TOOL / REGION FINGERPRINT

Will I be flagged as a Chinese user?

Websites and AI services (Claude, ChatGPT, Gemini…) infer your region from your network exit and browser characteristics, and may change how they treat you accordingly. This tool works from first principles: it checks all nine region signals a browser can expose, weighs them, and tells you which ones to fix.

Every check runs locally in your browser. Nothing is collected, nothing is stored, and the one server-side call only echoes your own network info back to you.

Probing your browser environment…

What a browser can actually expose

Region information splits into three layers. The network exit layer carries the most weight: the server sees your IP country directly, and no browser setting touches that; GFW blocking patterns further reveal whether you are sitting inside a mainland network. The system locale layer covers your language list, your timezone, and the most-overlooked one — the Intl default locale, which leaks your OS regional format through JavaScript even after you switch the browser language to English. The device fingerprint layer covers system-level censorship of the 🇹🇼 flag emoji, mainland-specific fonts (DengXian, FangSong_GB2312, MiSans and friends), Chinese browser user agents, and WebGL fingerprints from domestic GPUs such as Moore Threads.

Why no single signal is trustworthy

Chinese speakers abroad keep Chinese in their browser; Singaporeans live at UTC+8; some Linux users cannot render flag emoji at all. Every individual signal produces false positives. So this tool attaches a weight and a piece of raw evidence to each one, scores them together, and cross-checks consistency: an exit IP in the United States alongside a Shanghai system timezone is a proxy trace, and that mismatch can itself get you flagged.

What this tool cannot see

The country you entered when registering, the payment methods on file (a UnionPay card, a mainland phone number), and your historical login IPs all live in the provider's database, out of reach of any browser check. If your browser environment scores clean and you are still treated differently, the problem is almost certainly at the account layer, not the browser layer.

Concept credit: yArna/isChinaUser. This is an independent rewrite; the detection set has been extended to nine signals across three layers.