At a Glance

Imported a subscription but not sure which node to use? First make sure the client connects, then compare candidates on your usual network and with the tasks you actually do. For browsing and calls, prioritize consistent latency; for large downloads, check sustained speeds and the billing multiplier. Treat location as a hard requirement only when a service needs a specific exit region. Keep a comparable backup node, too.

Latency and Speed Tests: Check for Timeouts Before Comparing Results

A node is a set of server settings the client uses to route connections. The “latency” shown in a node list is the time taken by a particular test, not the response time of every website. A TCP test typically checks only whether the node’s port is reachable; a request made through the proxy also depends on the test URL and connection setup. Don’t rank results from these two types of tests together. Low latency doesn’t necessarily mean fast downloads, and fast downloads don’t guarantee clear, uninterrupted calls.

  1. In the v2rayN node list, select three candidates and choose “Test server real latency” from the right-click menu (wording may vary slightly by version). Use the same test and URL for each node, run five tests, and record timeouts as well as successful results.
  2. Keep the network and time of day consistent, and pause background syncing and downloads. Take the median of successful results and note the highest value too. Don’t record a timeout as 0 ms or rely only on the fastest run.
  3. Run another round during your usual evening hours. For browsing, compare the same set of frequently used pages. For calls, check jitter (latency variation) and packet loss (the share of data that never arrives) in the call app’s statistics. For downloads, track the average speed over time from the same source.
82 ms
Example: median time for successful requests on A
2/5
Example: timeouts on A
108 ms
Example: median time on B, with all five tests succeeding

In this example, A succeeds in 76, 82, and 91 ms, but times out twice. B’s five results range from 101 to 117 ms. For everyday browsing, try B first: avoiding occasional stalls is more useful than saving a little over 20 ms. These figures illustrate how to read results; they aren’t pass/fail thresholds. If all three nodes time out on the same URL, cross-check with another trusted test URL before blaming the nodes.

Traffic Multipliers: Estimate Plan Usage from Billing

A traffic multiplier is a factor providers use to calculate plan usage, not a speed boost. If a provider labels a node 2×, 1 GB of measured traffic will typically use 2 GB of your plan allowance. A multiplier in a node name is only a clue. The table below assumes the provider applies the multiplier directly to measured traffic, without rounding or other charges.

Node multiplierFor 10 GB of measured trafficUsable traffic with 30 GB remaining
0.5×5 GB deductedAbout 60 GB
1×10 GB deductedAbout 30 GB
2×20 GB deductedAbout 15 GB

Check the provider’s billing terms for the node first. Pause syncing on other devices, then note the remaining allowance before and after a test. Run a short test with a file of about 100 MB and compare the allowance after usage updates. Some plans count uploads or apply rounding or time-based rules, so don’t use file size to calculate an exact multiplier. If both nodes handle the task reliably, choose the one with the lower multiplier. Avoid repeatedly downloading large files just to compare peak speeds.

Choosing a Location: Distinguish the Node Entry Point from the Exit Location

The entry point is the server address the client connects to first; the exit is the proxy location websites see. With a relay node, these may be in different regions. “Hong Kong” or “Japan” in a node name might refer to the entry point, the exit, or simply a provider’s label. If you need a specific exit location, check the provider’s documentation, then check the exit IP’s region in a browser routed through the proxy. Trust what the service itself reports.

Choose a Location Based on Service Requirements

Everyday Browsing and Calls
  • Keep one candidate for each available location and test them on both your home broadband and your usual mobile network.
  • Before a call, run a short test in the actual meeting app and confirm it uses the intended route. A browser speed test can’t replace a call test.
  • For a backup, choose a different route rather than two similarly named nodes with the same entry point.
Services Requiring a Specific Location
  • First choose exit locations that meet the service’s requirements, then compare stability and multipliers among those candidates.
  • Verify the location in the same browser or app you’ll use. If you use split tunneling, make sure the site being tested isn’t matching a direct-connection rule.
  • Keep the primary and backup nodes in the required region where possible to minimize the impact of switching routes on login sessions.

If you don’t need a specific location, choose based on real-world performance. If you do, treat the exit location as a filter—not a speed rating inferred from the node name.

Protocol Types: Check the Core and Server Configuration

VMess and VLESS are proxy protocols; WebSocket is a transport; TLS and REALITY are connection security settings. They aren’t interchangeable options in a single list. “VLESS + WebSocket + TLS” and “VLESS + TCP + REALITY” both use VLESS, but require different settings. The core is the part of the client that establishes the proxy connection. Being able to import a node doesn’t mean the current core supports its full configuration.

Stick with a Configuration You’ve Verified

Recommended

Import the provider’s configuration from your subscription, keeping the protocol, transport, and security settings intact. If nodes in the same location with similar multipliers all work, compare them against real tasks at the same time of day. Don’t guess speed based on protocol names.

Best for: Choosing a primary node when you already have working options

Choose a Compatible Option for Your Core

On desktop, v2rayN uses the core assigned to the node. Android v2rayNG uses the Xray core, while v2flyNG uses the v2fly core. For a VLESS + REALITY node, check that your Xray version supports it and that you have the parameters provided by the server. With the v2fly core, choose a configuration the provider explicitly marks as compatible, such as a supported VMess node.

Best for: A subscription imports on two devices, but only one can connect

If only newly imported nodes fail, update the relevant subscription group first, then check the node’s transport, security settings, and port. For example, if the server specifies port 443, keep 443—but don’t assume the port alone means TLS is enabled. If the logs report an unsupported setting, resolve the compatibility issue. Don’t manually change VMess to VLESS or disable certificate verification just to make a test pass.

Common Questions About Choosing Nodes

Why does video buffer when latency is only 60 ms?

Latency measures how quickly one request completes; video also needs sustained throughput. At the same time of day and the same quality, play the video on each node for three minutes and count the buffering events. If only that video service stutters, test another trusted download source to tell whether the node is generally slow or the route to that service is the problem.

Why does my phone disconnect when I switch networks, even though it works on my computer?

Test the same node on your phone over Wi-Fi and mobile data, and note when each failure occurs and what error appears. If only mobile data is unstable, choose a backup node while on that network. If both networks fail, check that the phone’s client version, core, and subscription configuration are compatible with the setup on your computer.

The node test succeeds, but websites won’t open in my browser. Should I switch locations?

First, open the v2rayN tray menu and check “System proxy.” Make sure “Automatically configure system proxy” is selected, then check whether the browser uses the system proxy and whether routing rules send the site directly. If another site works but the one you want still doesn’t, try a backup node with the same URL. There’s no need to reshuffle every location first.

Both low- and high-multiplier nodes connect. Which one should I keep?

Run the same everyday task on both nodes during the evening. If both finish within an acceptable time, set the lower-multiplier node as your primary. If the higher-multiplier route is more reliable for evening calls, keep it for calls. Note the node name, multiplier, and test time so you have a baseline for future comparisons.