From Project V to core families
Project V is an open-source ecosystem built around network proxy technology, with V2Ray among its best-known project names. As the ecosystem evolved, V2Fly continued community maintenance of related cores, while Xray developed independently on shared technical foundations. The projects have historical ties and share some protocols and configuration concepts, but their feature sets and specific fields do not always match.
A search for “V2Ray” often turns up protocols, cores, clients, and subscription services together. They are not the same thing. A client provides the interface, a core runs the configuration, a server provides the connection endpoint, and a subscription distributes node details in bulk. Knowing how these layers differ helps pinpoint whether an issue lies in the interface, configuration, system network, or server.
That’s why a recognized share link does not guarantee that every setting suits the current runtime environment. If a protocol or transport is incompatible, check the core’s capabilities against the server requirements rather than relying on the vague claim that it “supports V2Ray.”
Xray core and client selection
v2rayN is a graphical client for desktop platforms. It manages nodes, updates subscriptions, integrates with the system, and can call compatible cores based on its own support. v2rayNG is for Android and uses the Xray core; v2flyNG is an Android alternative for the v2fly core. Choose based first on platform and configuration requirements, then on your preferred workflow.
Core names are not a ranking of connection quality. Real-world performance also depends on server status, transport settings, the DNS resolution path, and your local network. If the same node behaves differently in two clients, compare the active node, routing mode, DNS, and traffic-capture scope first. Local setup differences can easily be mistaken for core performance differences.
To compare the three clients, start with the client comparison. If your configuration already works and there’s no specific compatibility issue, keep the current environment reproducible. A targeted migration is easier to verify than changing clients, cores, and rule sets all at once.
Open-source licenses and usage
All three clients are open-source software and can be used for free. The projects use open-source licenses; permitted use, modification, and redistribution are governed by each project’s license and the license texts for included components. The graphical client and core are separate software components, so one project’s license does not automatically apply to every component.
Open source describes the code and its licensing; it doesn’t choose a subscription service for you or mean every network endpoint is trustworthy. Keep subscription URLs, node credentials, and backup files secure. Before sharing logs, remove identifiable server addresses, user IDs, and access credentials. Keep only the error type and relevant settings needed to diagnose the issue.
This site focuses on download links and Chinese-language setup guides; it does not distribute nodes or host subscriptions. Whether a client connects successfully depends on a valid configuration, a compatible server, and local network conditions. Installation and network access are separate steps that need to be verified.
Update the client, core, and rules separately
Client, core, subscription, and rule-resource updates each affect different things. A client update may change menus, system integration, or how configurations are generated. A core update affects protocol and runtime behavior; a subscription update replaces node details; a rule-resource update changes classification and matching. Updating one does not mean the others have changed.
Before making changes, back up your current configuration and note the active node and routing mode. After updating, retest a scenario that worked before enabling new features. If something goes wrong, keep the logs and revert settings one at a time. That makes the cause easier to find than clearing subscriptions, reinstalling the client, and changing DNS all at once.
Menus and default behavior can change between releases, so these guides focus on what settings do and the order of operations. Choose an installer on the download page. For platform-specific preparation, permissions, and configuration checks, see the complete installation and setup guide for all platforms.
Initial setup: validation order
First, confirm that updating the subscription actually added selectable nodes, then choose an active server manually. Next, check that the core starts, and only then test the target app. If the browser works but other apps fail, check the traffic-capture scope. If the logs show a DNS resolution error, trace the DNS path. If no apps can connect, check for leftover system proxy settings, TUN routes, and local network issues. Change one variable at a time so you can compare results.
V2Ray node speed tests can help narrow down options, but they don’t replace testing real access. Test results depend on the request address, protocol, and timeout settings. One timeout doesn’t prove a node is permanently down, and one successful test doesn’t prove every app will work. Compare nodes under the same conditions, then test the services you actually need. Consider V2Ray subscription conversion only when the target client requires a different format. Conversion can omit fields or expose the full subscription URL, so it shouldn’t be the default fix for an import failure.
System proxy is not a routing mode, and TUN does not mean “proxy everything.” The former determines how traffic enters the client; rules still determine where it goes. V2Ray DNS settings also depend on how traffic is captured: app-level, system, and core-level resolution may follow different paths. Start with a minimal working setup, then add rules, sniffing, or FakeDNS as needed to reduce misdiagnoses caused by interacting features.
Start with subscriptions in the v2rayN setup guide →