How We Differ from Other VPNs

We are a service **optimised for the mainland Chinese network**, not a reseller of international lines. The international exits of China Telecom, Unicom and Mobile differ enormously, so we tune per carrier and recommend different regions to each; we have landing nodes inside China so users on mediocre broadband still have a stable entry; and we keep adjusting ports and protocols as the domestic network changes rather than setting up once and walking away.

In the other direction, overseas Chinese reaching Chinese services simply choose the China region on the same account: Chinese sites leave from a Chinese IP while overseas sites still exit abroad. Reaching China from abroad and reaching abroad from China are the same thing in two directions.

• Their strengths: dozens of countries, thousands of servers, polished apps, independently audited no-log claims. If you live in Europe or America and want encryption on public Wi‑Fi or another country’s streaming catalogue, they are a fine choice.
• Their weakness: the product was not designed for China’s network. A proprietary protocol has a fixed fingerprint, and once recognised inside China the whole app stops working; with no landing nodes in China, the “China” feature either does not exist or uses Hong Kong or Taiwan IPs that Tencent Video and iQIYI still block.
• Our approach: standard protocols only — AnyConnect, OpenVPN, sing-box, Tailscale and an HTTPS web proxy — with clients from the vendor or the open-source community. What gets blocked is a single address, replaced with a new one; the client can never be pulled from a store.
• Our trade-offs: 24 regions and far fewer servers than the brands; no flashy app, so setup takes a few more taps (QR codes and one-click imports keep that as short as we can); no third-party audit report — what we log and never do is written on the About page, and we rely on stating it clearly rather than on a stamp.

• Their strengths: cheap, many nodes, a subscription link you paste and go; people who enjoy tinkering with clients like the freedom.
• Their weakness: most are individuals or small teams renting overseas machines and reselling traffic, with no landing nodes of their own inside China, so exit quality moves with whatever they rent; disappearing, renaming and shutting down are common, and a paid year may not be served in full; clients are entirely your own problem and nobody is accountable when things break.
• Our approach: the machines are ours, and the mainland exits are our own identity-verified machines at a mainland cloud provider; one account gives six connection methods, a whole-home router option and 24-hour support groups that airports do not have; we have operated since 2007, and accounts, expiry dates and prices did not change through the rebrand.
• Our trade-offs: a higher price than an airport (US$7 a month for a standard account); no list of hundreds of nodes to pick from — each region has several machines and the system picks the least busy one, so you choose a region, not a server.

• Their strength: they cost nothing, and for an occasional lookup that is enough.
• Their weakness: mainland exit bandwidth is metered, so free can only survive by overcrowding, capping or selling data; browser-extension proxies carry only web requests, not video players, apps or system DNS; and an installer of unknown origin is a risk in itself — see the guide on recognising risky VPN apps.
• Our approach: a 24-hour free trial with no credit card, for testing; pay-per-day points for occasional use; and 30 days added for every friend who pays, a lasting source of free time.
• Our trade-off: there is no permanent free tier. The cost structure does not allow one and we do not pretend otherwise.

The private network (Tailscale) logs in once and stays online with region switching in the menu; Cisco AnyConnect has the best compatibility with built-in or official clients; OpenVPN serves routers, NAS and Linux; the web proxy routes only the browser and never drops; sing-box resists packet loss and is the fastest; the router option covers TVs, boxes and grandparents’ devices for the whole home.

No single method is best on every network, so we never lock you into one: switch any time on the same account — from sing-box to AnyConnect when a campus network throttles UDP, to the web proxy when broadband is unstable. How to choose is covered in the connection-method guide.

• 5 devices online at once, the same for standard and premium; install on as many as you like — only simultaneous connections are limited, and the longest-idle device is dropped when you exceed it, so a new phone does not require deleting the old one.
• Standard and premium switch freely with remaining time converted at equal value (premium costs twice as much); router access is premium-only, which is how we price the router as a whole-household option — a single user does not need premium.
• Refunds are pro-rata for time used and available any time, handled by a person; we do not promise a settlement date.
• sing-box and OpenVPN profiles carry your credentials, and a password change revokes old profiles — a safety design, which also means re-importing after you change your password.

• We are not an anonymity service. We keep what billing needs — email, plan, expiry, and the duration and volume of each connection — and never log content, inject ads or sell data; client IPs never leave our servers. But “no content logs” is not “anonymous”, and the About page says so plainly.
• We do not build an app. Clients come only from the official upstreams — Cisco, OpenVPN, sing-box/Hiddify, Tailscale — mirrored unmodified, so “our app was removed from the store” can never happen; the price is one more setup step than a one-tap app.
• We do not promise speed figures. Cross-border bandwidth depends on the carrier’s exit, the hour and your own broadband; everyone is slow at the 8–11 pm national peak. We offer ways to change region and protocol, not a guaranteed number.
• We do not pass off Hong Kong or Taiwan nodes as China. The China region is mainland machines; Hong Kong was permanently retired in August 2026 as its local network deteriorated, and it will not be replaced with HK or TW IPs.