This article explains how to import nodes into sing-box, handle incompatible subscriptions, and reduce frequent changes to network egress when using it as a team. First confirm which import formats your client supports, then add the configuration and test the connection. A successful node import does not necessarily mean the account environment will be stable.
Before importing: distinguish node links, subscriptions, and configuration files
sing-box is a proxy core, and the menus and import capabilities vary across graphical clients. Support for the sing-box core does not mean a client can directly import every node link or Clash subscription. Before proceeding, check the client documentation to confirm which input formats it supports:
- Single-node links: commonly beginning with vless://, vmess://, or trojan://; the client needs to support parsing them.
- Subscription URLs: used to retrieve configurations or node lists; the returned format must be compatible with the client.
- sing-box JSON configurations: usually include content such as outbounds and routing. A snippet containing only node parameters may not be runnable on its own.
When obtaining free nodes from sources such as freevpn-x.com, also check the protocol, expiration date, and format. Don’t assume a file is compatible based only on its extension.
Import nodes using a method supported by your client
- Install the client from the project’s official release page or a trusted app store, and note its version. Teams should try to use the same client and version to minimize differences in configuration behavior.
- Open the configuration or profiles page and look for an option to add, import, or subscribe. The exact label depends on the client version.
- If clipboard import is supported, copy and paste the complete link. For a subscription, enter the full URL and refresh it. Clients that support remote configurations may also require the URL to return sing-box JSON directly.
- If you have only a local JSON file, use the file import feature. Users of the command-line core must provide a complete configuration; a node link cannot be used as a configuration file.
- Save and select the configuration, grant VPN or network extension permissions if prompted, start the connection, and then check that commonly used websites load.
Do not submit subscription URLs containing tokens to untrusted online conversion services. If the format is incompatible, first ask the provider for a compatible configuration, or use a conversion method explicitly supported by the client.
Team use: how node imports relate to account stability
The import method itself does not determine whether an account will be flagged for verification, but the selected egress after import can affect the usage environment. Automatic speed tests, failover, and subscription updates may change the egress IP or region. Teams should specify the configuration, egress region, and update method each member should use, so problems can be traced if someone switches settings unexpectedly.
A fixed node name does not mean a fixed egress IP. Providers may change routes, and free nodes may stop working. Before important operations, check the current egress IP and region. Keep devices, browsers, and sign-in methods reasonably consistent, enable two-factor authentication, and follow the account platform’s rules for team use. These steps cannot guarantee that you will avoid risk controls.
If the import fails or you can’t connect, troubleshoot in order
If you see a parsing error, first check that you copied the entire link, the subscription hasn’t expired, and the returned content isn’t actually a login page. If you see a configuration field error, check whether the configuration is compatible with the core version; don’t delete fields blindly.
If the configuration starts but you can’t access anything, first turn off the proxy to confirm that your local network is working. Then check the logs for connection, authentication, or DNS errors, and verify the system time and VPN permissions. If a node is down, test another working node. Avoid switching nodes repeatedly in batches immediately before important account operations. Before sharing logs for troubleshooting, redact subscription tokens, passwords, and UUIDs.