Back to BlogThe EDIS Global Free Latency Test compares browser response times to VPS locations around the world. Use it to shortlist locations for SSH, remote desktops, VPN access or other interactive applications before you order.
Run the free VPS latency test. No signup or email is required.
The results describe the connection from the device and network running the test. They do not measure download speed, VPS processing power or the experience of every customer who might use your service.
How to compare VPS locations
- Use your normal connection. Open the test on the connection you intend to use. Pause large downloads and uploads. If your setup includes a VPN, test through that VPN as well as directly.
- Start the test. Choose Start latency test. Progress shows how many available locations have finished. You can stop at any time; completed measurements remain visible.
- Narrow the results. Use Search locations and Region to focus on suitable locations. These filters change what you see; a new run still tests all available locations.
- Choose a comparison metric. Select Minimum latency or Average latency under Compare using, then Lowest latency under Sort by. Recommendations use your selected metric. Locations sharing one test endpoint appear once in the recommendation list.
- Review your shortlist. Repeat the test when useful. Use View VPS plans or Looking Glass to investigate a location before ordering.
Minimum, average and maximum latency
Latency is the time a request takes to travel to a test endpoint and back, shown in milliseconds (ms). Lower values mean a quicker round trip.
- Minimum: the quickest measured response. It is useful for comparing the lowest observed delay, but one quick response does not describe every exchange.
- Average: the mean of the measured responses. It gives more weight to slower samples than the minimum does, so it can change the order of your shortlist.
- Maximum: the slowest measured response. A large gap between the minimum and maximum is a reason to repeat the test and investigate variation.
The tool collects ten samples after an initial warm-up exchange. A partial result is labelled separately. Only completed results enter the lowest-latency recommendations.
What the latency ranges mean
The current tool uses four descriptive ranges. They apply to the minimum or average value selected under Compare using.
| Rating | Measured response time |
|---|---|
| Low latency | 50 ms or less |
| Moderate latency | Above 50 ms, up to 100 ms |
| Higher latency | Above 100 ms, up to 250 ms |
| High latency | Above 250 ms |
These ranges are a comparison guide, not a guarantee of application performance. An interactive remote desktop can be sensitive to delay, while a background transfer may be less sensitive. Packet loss, available bandwidth and the application itself also affect the experience.
Choose for your users and your network
The geographically closest location is not always the one with the lowest latency. Your internet provider's routing, peering, congestion, Wi-Fi and any VPN in the path can affect the result.
If you are choosing a VPS for your own remote access, test from your usual connection. If you are deploying a service for customers, also test from networks in the regions where those customers connect. A result from your laptop cannot establish their latency.
Repeat the comparison at different times before making a decision based on a small difference. The tool reports a short sample of conditions at the time of testing; it does not establish long-term reliability or calculate a dedicated jitter or packet-loss metric.
Use the location's Looking Glass to investigate connectivity further. Location pages also provide information to help you compare available VPS plans.
For more background on the network, read why EDIS Global.
Stopping, retrying and saving results
Stop test ends the current run. Run again starts a fresh comparison. After a run finishes or stops, Retry is available for a location whose test failed or was stopped. Locations sharing the same endpoint receive the same retried measurement.
Copy results creates a text summary with the test time, sample counts, result states and measurements. Results can also be restored in the same browser tab for up to 30 minutes, where browser storage is available. Restored results are labelled and do not start a new test automatically.
A No response or Incomplete test message means that the browser test could not complete. It does not prove that the VPS location is offline. Browser settings, a firewall, an unavailable test endpoint or a connection problem may affect the test.
A Test unavailable location has no configured browser test endpoint. It stays visible so you can open its Looking Glass or VPS plans instead. The page shows the current location count and how many are available to test.
How the browser test works
The tool uses secure WebSocket connections to configured Meter.net endpoints. It measures repeated round trips after a warm-up exchange and limits how long it waits for a connection or response.
Browser WebSocket results can differ from command-line ICMP ping because they use different protocols and processing paths. Compare locations within the same test method and from the same connection, then validate the application you plan to run.
Frequently Asked Questions
A good result depends on your application and the networks involved. The tool labels 50 ms or less as low latency, above 50 through 100 ms as moderate, above 100 through 250 ms as higher, and above 250 ms as high latency. These are comparison ranges, not a guarantee of smooth performance. Compare both minimum and average results, repeat the test and try your actual application.
Ready to compare? Start the free latency test and review the locations that respond best from your connection.