· Node Selection · About 8 minutes

How to Choose a V2Ray Node: Latency, Traffic Multipliers, Locations, and Protocols

Not sure which node to pick from your subscription? Learn how to compare latency tests, traffic multipliers, server locations, and VMess, VLESS, and Trojan protocols.

At a glance

If you’ve imported a subscription but aren’t sure which node to use, first rule out nodes with incompatible cores or that can’t reach the services you need. Then compare real connection latency and failure rates on the same network, and factor in the traffic multiplier. For everyday use, prioritize stability and data usage. Choose a backup on a different route so an outage on one route won’t take out both.

Understanding latency: test real connections, then try your actual use cases

In v2rayN, select a candidate in the node list, right-click, and choose “Test server real connection latency.” If the menu also offers a TCP latency test, use it as an initial check that the endpoint is reachable. A successful TCP connection to the configured address and port doesn’t mean the protocol handshake will succeed or that the node can reach your target service. A real connection test tries to access a test address through the node, so its result is closer to how proxy requests respond—but it’s still just one measurement to that test target, not a download speed test.

  1. Start with three candidates. Use the same device, network, and test address, and test each node five times. Record successful response times and timeouts. Don’t switch Wi-Fi or run large downloads during the tests.
  2. Compare the median of successful results separately from the number of failures. For example, node A succeeds three times at 72, 78, and 85 ms, but times out twice. Node B succeeds all five times, with response times between 108 and 135 ms. Even though A is faster when it connects, B is the better everyday choice.
  3. Switch to a candidate and make sure the test app’s requests are actually routed through the proxy, then open sites you use regularly. If downloads are your main use, compare sustained speeds by downloading the same file from the same source, one node at a time. If you often join meetings, test a call during your usual hours and watch for freezes, disconnects, and reconnections.

The figures below illustrate the example above; they aren’t measured results or recommended latency thresholds. Results from different test addresses, carriers, or times of day shouldn’t be ranked together.

5 tests
Tests per candidate
78 ms
Median successful response time for example A
2 timeouts
Timeouts for example A

How traffic multipliers work: check your provider’s billing rules

Labels such as “0.5×,” “1×,” and “2×” in node names usually indicate how the provider calculates plan usage. They aren’t v2rayN settings or speed ratings. Before comparing nodes, check whether your plan counts downloads, uploads, or both, and whether the multiplier varies by time of day. The client’s traffic stats can help you track usage trends, but the provider’s account balance determines how much data is deducted.

Example of quota deductions based on metered traffic multiplied by the node’s multiplier
Node multiplierMetered trafficQuota deducted
0.5×10 GB5 GB
1×10 GB10 GB
2×10 GB20 GB

Suppose the 10 GB here refers to downloads, and the provider also counts 2 GB of uploads. A 2× node billed on combined upload and download traffic would deduct 24 GB. If the billing rules aren’t clear, note your account balance, pause other devices using the same subscription, then generate a small, identifiable amount of traffic on just this node and check the balance after it updates. Don’t repeatedly run download speed tests across every node just to compare multipliers.

Choosing a node location: distinguish the entry point, exit location, and actual route

The location in a node’s name is just the provider’s label. The entry point is the address your device connects to; the exit is the public IP address seen by the destination website. With a relay, they may be in different locations. A shorter distance doesn’t guarantee lower latency: peering between networks, indirect routing, and peak-hour congestion can make a nearby candidate slower.

A stable, low-multiplier node on your usual network

Recommended

First make sure the sites you use regularly load, then retest during your usual hours. If two nodes both meet your needs, choose the one with fewer timeouts and a lower multiplier—not the one with the more appealing location label or the fastest single result.

Best for: everyday browsing, document syncing, and frequently used services.

An exit in a specific location that can access your target service

Check the exit location in a browser that you’ve confirmed is using the proxy, then test the actual service you need. IP geolocation databases can disagree, and the location shown by a lookup site doesn’t guarantee that a target service will see it the same way.

Best for: tasks that require an exit in a specific location.

A backup node on a different route

If the provider identifies different entry points or routes, choose a backup on another route. Without route information, choosing a different location broadens your options but doesn’t guarantee it won’t fail at the same time as your primary node.

Best for: switching when the primary route is congested or temporarily down.

VMess, VLESS, and Trojan: check core compatibility first

The protocol determines how the client connects to the server; it doesn’t directly determine which route is faster. The same protocol can use different transport and security settings. If a connection fails after import, check the protocol, transport, and security settings first. Don’t simply label a node that won’t connect as high-latency.

ProtocolWhat to check before connecting
VMessCheck the UUID, transport, and security settings required by the server. If only nodes of this type are failing, also check that your device’s date and time are synchronized.
VLESSCheck the UUID, transport, and security type. If a node specifies REALITY or particular flow-control parameters, the core you’re using must support that combination.
TrojanCheck the password and security settings. For entries using TLS, also check the server name and certificate verification result. A certificate error isn’t a latency issue.

On desktop, v2rayN uses the core configured for that node. On Android, v2rayNG uses the Xray core, while v2flyNG uses the v2fly core. Being able to import the same subscription doesn’t mean every node will work in both clients. If a configuration isn’t supported, choose a compatible entry from the subscription or use a desktop core that supports that combination. Don’t just change the protocol from VLESS to VMess and try again.

Keep separate primary and backup nodes for each network

The top pick on your home broadband might not work as well on mobile data. Changing networks can affect endpoint reachability, route congestion, and exit performance. If you regularly use two networks, run a small test on each; it’s more reliable than reusing an old latency ranking.

Two sets of candidates on the same subscription

Desktop broadband

  • In v2rayN, test three nodes first. Set the one with few timeouts that can complete your task as the primary node.
  • Choose a backup on a different route. Test it during your usual work or meeting hours, not just during off-peak daytime hours.

Android mobile data

  • Turn off Wi-Fi before testing. Use v2rayNG or v2flyNG with a core that supports the node configuration to check the apps you use regularly.
  • If per-app proxying is enabled, confirm that the test app is included. Prioritize checking connectivity and page loads, and keep large-file tests to a minimum.

Record the network, time of day, failures, multiplier, and results for your target services. If you change carriers or your subscription updates its routes, you only need to retest your primary and backup first.

Node testing and switching: common questions

Why are downloads still slow when real connection latency is only a few dozen milliseconds?

Latency measures how long a test request takes to get a response; throughput measures how much data a connection can transfer over time. Pause other transfers, then download the same file from the same source over two nodes during the same time period. Compare speeds once they stabilize. If only one source is slow, try another source to avoid mistaking a server-side speed limit for a node problem.

Every node in my subscription times out. Does that mean the entire route is down?

First, compare with a known-good configuration and check the core logs for startup, handshake, or parsing errors. If only one test address fails, open “Settings” → “Parameter Settings” in v2rayN and check whether your version lets you configure the test address. Switch to a test address you can verify, then retest the candidates under the same conditions. Don’t compare latency results from before and after changing the test address.

I switched nodes, but why does my browser still show the same exit IP?

First confirm that the page request is actually routed through v2rayN. Then check whether the current routing rules send the lookup domain directly. If the request is proxied, close the existing browser session and reopen the lookup page. Two nodes may also share the same exit IP, so an unchanged IP alone doesn’t prove the switch failed.

A familiar node disappeared after I updated my subscription. Can I find it by its old name?

First confirm that you updated the right subscription group. The display name may have changed, or the provider may have removed the entry. Don’t assume it’s the same route based on the name alone; check protocol compatibility and exit location against the new list, then briefly retest your primary and backup candidates.

Client downloads