Before You Begin: Install the Client and Get a Subscription
Choose a client from the Windows download page, install it, and open the main window. You don’t need to edit YAML or install service mode before your first launch. Once the main window appears, make sure you can switch between pages. If the client reports that the core failed to start, resolve that issue before importing a subscription. Installing the client only gets the app ready; you still need a valid configuration to choose a proxy node.
Get a subscription URL compatible with Clash or mihomo. You’ll usually find it on your subscription provider’s account page under an option like “Copy subscription URL”—not in the browser’s address bar. The URL may contain personal access credentials, so paste it only into a trusted client. Don’t share it in public groups, screenshots, or online conversion tools. This site provides downloads and configuration guides; it doesn’t issue subscription accounts or generate working nodes.
Before you begin, make sure regular websites load directly, then close any other proxy tools or VPN clients for now. If multiple tools manage your system network at once, it’s harder to tell what’s happening during the connection test. Don’t enable TUN, change DNS, or add custom rules yet. Complete the four steps one at a time; if something goes wrong, stop there rather than layering separate issues on top of each other.
Step 1: Import a Subscription and Activate the Configuration
Open Subscriptions and paste the full URL
In Clash Verge Rev, click Subscriptions in the left sidebar and find the subscription URL field at the top. Paste the full URL you copied. Check that it starts with https:// or another scheme explicitly provided by your subscription provider, that the end hasn’t been cut off, and that no extra text is included. Click Import and wait for it to finish. Button layouts vary by version, so use the page’s Import action. Don’t paste the URL into the proxy-group search box.
After a successful import, a configuration card appears on the Subscriptions page, usually with a name and related actions. Click the card or use its activation control to make it the current configuration. Seeing the card alone isn’t enough: open the Proxies page and confirm that proxy groups and their options are listed. If you have several subscription cards, activate only the one you need so you don’t mistake nodes from an older configuration for the new subscription.
Some configurations show subscription details such as remaining traffic or expiration date, depending on what the provider returns. These details don’t prove that the connection works, and their absence isn’t a reason to import again. For this step, confirm that the configuration is active, the Proxies page has selectable policies, and the client reports no configuration parsing errors. Then choose a proxy mode and node.
Import failed? Check the URL and network separately
If the page reports a download failure, first confirm on your provider’s account page that the subscription is still active, then copy the URL again. The account login page, plan purchase page, and subscription URL are separate links. If the logs show a timeout, check whether your current network can reach the subscription server. If parsing fails, make sure you copied a Clash- or mihomo-compatible configuration, not one made exclusively for another client. Don’t click Import repeatedly and create duplicate cards.
A subscription URL opening in a browser doesn’t guarantee that it contains a valid configuration; the server may return a login page or an error message. Never post a full response containing personal credentials in a public forum. Note the error type, the step where it occurred, and the client name, then check the Help Center. Importing local YAML, subscription overrides, and merging configurations are advanced topics. For your first connection, follow the subscription workflow here.
Step 2: Choose Rule Mode and a Proxy Group
Switch to Rule mode
Once the subscription is active, open the Proxies page, find the mode selector, and choose Rule. In Rule mode, the current configuration’s rules decide whether a connection goes direct or through a proxy; not every site has to use the same exit. Start with this mode, and don’t leave Global enabled just to make troubleshooting easier—it can also change the route for traffic that should go direct.
Three modes to know: Rule routes traffic according to the configuration; Global sends connections to the global policy, which still needs a usable policy selection; and Direct usually bypasses proxy nodes. The mode controls how the core handles traffic it receives. It doesn’t make your browser send requests to the client automatically. After selecting Rule mode, enable the system proxy in the next step.
Choose a node in the policy group you’ll use
On the same page, find the policy group responsible for proxied traffic, expand it, and select an available node. Group names come from the subscription configuration and may be “Node Selection,” “Proxy,” or something else, so don’t rely on a specific name. If the group contains a nested automatic-selection group, choosing it lets that group pick the node. For easier troubleshooting, you can instead select a specific node in a group that allows manual selection.
Latency tests can help narrow down your options, but a successful test doesn’t mean every website will load, and a timeout doesn’t necessarily mean the node is unusable. Test URLs, protocols, and real app requests may differ. Choose a candidate node and keep it selected for the connection test. If it fails, try another node in the same group and compare. Don’t change the configuration, mode, and node all at once.
If the Proxies page shows only Direct or none of the expected nodes, go back and check that the configuration is actually active. Don’t use Global to hide a configuration that hasn’t loaded. For rule order, policy group types, and domain-based routing, see the User Guide. For now, confirm that Rule mode is selected and the policy group for proxied traffic has a clear selection.
Step 3: Enable the System Proxy and Connect
Route browser traffic through the client
Open Settings, find System Proxy, and turn it on. Some versions also have the same toggle in the main window or tray menu. This changes the current user’s system proxy settings so apps that follow them send requests to the client’s local proxy port. For your first test, leave TUN off and use one clear traffic path so you can pinpoint where a problem occurs.
The system proxy and proxy node are separate settings. The local address and port in the system proxy point to the client running on your computer; the subscription’s node is the remote exit the client uses to forward requests. Don’t enter the node server’s address in Windows’ proxy server field, and don’t force a port from another guide. Clash Verge Rev commonly uses 7897 as its mixed port, but check the value in your client’s settings.
To check the settings in Windows 11, open Settings → Network & internet → Proxy and look at the address and port under Manual proxy setup. The page names are similar in Windows 10. Don’t enter different ports in Windows and the client. If another program is using the port, close the conflicting tool or resolve the client’s startup warning before turning on the system proxy again.
Keep the core running, then open a website
Make sure the client isn’t repeatedly reporting a core startup failure or port conflict. Then open a browser such as Microsoft Edge, which typically follows the Windows system proxy, and visit a site expected to use the proxy. Keep the client running; minimizing it to the tray is fine, but don’t exit the process. Wait for the page to load. If you see a blank page or an error, note the browser message and check the connection logs in the next step.
The system proxy doesn’t route every app. Browser extensions, Firefox’s separate proxy settings, and some command-line tools may use a different network path. If you use a proxy extension, disable it for the test or confirm that it uses the system settings. Don’t assume the whole setup is broken just because an app that ignores the system proxy behaves the same. Get one regular browser working first, then troubleshoot other apps separately.
Step 4: Verify Access and Routing Rules
Check the website and connection logs
First, open a website you can normally access directly to make sure the proxy settings haven’t disrupted your internet connection. Then visit a site expected to use the proxy and confirm it loads. A blue toggle or a tray icon doesn’t prove that a request went through the selected node. In Rule mode, the two sites may take different routes—that’s normal. They don’t have to show the same exit.
Open the Connections page in the client while the test page loads and check for a connection to its domain. Expand the connection details to see the matched rule, policy, or proxy chain; the exact field names vary by version. For a site expected to use the proxy, you should see the expected policy group or node. If the rule specifies a direct connection, you may see DIRECT. This distinguishes “the page loads” from “the request took the expected route” and prevents mistaking a successful direct connection for a working proxy.
The Connections page may show only active connections, so a short-lived request might not stay visible. If you can’t find the destination, leave the page open and reload the test site, then check the Logs page for entries at the same time. Browser caching can also reduce new requests, so don’t draw conclusions from refreshing a fully cached static page. The level of detail in the logs depends on the log level; you don’t need to enable the most verbose logging for a first test.
Go back to the step that matches what you see
No browser requests appear: Check that the system proxy is still on, the client is running, and the browser uses the system proxy. Another proxy tool may have changed the system settings. Focus first on whether traffic is reaching the client. Don’t change node protocols or DNS settings before the core has received the request.
A request appears, but the destination connection fails: Check whether the connection details show the policy group you just selected, then try another candidate node on the same site. If switching nodes fixes it, the original node or its route to the site is the more likely cause. If every candidate fails, check the logs for timeouts, connection refusals, or DNS errors, and confirm the subscription is active. One site failing doesn’t mean the entire subscription is unusable.
The browser works, but one app doesn’t: Check whether the app supports the system proxy, has its own proxy settings, or needs to restart before it picks up system settings. Windows Store apps may also be affected by loopback access restrictions. Diagnose these cases separately instead of switching straight to Global. See the Help Center for app routing and connection troubleshooting, then decide whether to use TUN or a loopback exemption.
Keep the working settings and exit in order
Once the connection is verified, note which subscription is active, that Rule mode is selected, and which policy group you chose. When updating the subscription, update the existing card instead of deleting and recreating it each day. After an update, quickly check that the policy group selection is still appropriate. If the connection drops after a network change or waking from sleep, check the system proxy and node status before reinstalling the client.
Before fully closing the client, turn off the system proxy, exit the process, and confirm that regular websites still load. If the browser says the proxy server is unavailable after an unexpected exit, check Windows’ manual proxy settings and disable any proxy entry still pointing to the stopped client. You’ve completed the first-time setup: the subscription is active, Rule mode is selected, browser requests reach the client, and the connection details match the expected route.
Continue with the guide that fits your needs: for configuration structure, rule-based routing, and TUN, read the Clash User Guide; for a specific error, visit the Help Center. Keep this verified baseline and change one setting at a time so you can quickly identify what caused a problem.