Why a VPN can change speed and ping
Without a VPN, your device sends traffic through its usual route to an internet service. With a VPN, traffic travels through an encrypted tunnel to a VPN server before reaching its destination. The extra hop may increase path length and processing, although some routes can behave differently depending on the internet provider and destination. The outcome is not determined by encryption alone.
A reproducible six-run comparison
- Choose a consistent test. Use the same computer, Wi-Fi/Ethernet connection and speed-test endpoint. Pause cloud backups and other heavy background transfers.
- VPN off: Run download, upload and latency tests three times, with short breaks. Note the individual results, not only the best one.
- VPN on: Connect to a specified VPN server, record its country and protocol, and run three more tests using the same endpoint.
- Compare medians. Find the middle result in each three-test group. This reduces the impact of a single unusually good or bad run.
- Repeat another time. Network conditions change. If results are close, alternating off/on runs on a second occasion helps check whether the difference is persistent.
- Change one VPN setting. Try a nearer server or another supported protocol, then repeat the same comparison before drawing conclusions.
Calculate speed retained and speed lost
Speed retention (%) = VPN-on download speed ÷ VPN-off download speed × 100. Speed loss (%) = (1 − VPN-on speed ÷ VPN-off speed) × 100. Using the same formula for upload speed is possible, but report it separately from download.
| Worked example (illustrative) | Value |
|---|---|
| Median download, VPN off | 200 Mbps |
| Median download, VPN on | 140 Mbps |
| Speed retained | 70% |
| Speed reduction | 30% |
| Idle latency off / on | 18 ms / 42 ms |
| Latency increase | 24 ms |
These values illustrate arithmetic only; they are not live NextWatch AI readings or typical results promised by a VPN brand. The VPN Speed Test can help you compare your actual results, including speed retained and latency differences.
Practical ways to improve VPN performance
Start with geography: a VPN server closer to you often reduces unnecessary travel, though overloaded local servers may still perform poorly. Check load: if your app exposes server capacity, compare available endpoints. Try a supported protocol: the performance of WireGuard-based, OpenVPN and other tunnel modes depends on hardware, network filtering and provider implementation. Do not assume one protocol is always fastest.
Compare connection types: poor local Wi-Fi can dominate the result whether the VPN is on or off. Review background apps: encryption does not create extra capacity; simultaneous downloads can still saturate the connection. Check security features: additional filtering or multi-hop routes may change speed, but disabling protections changes privacy and security tradeoffs. Do not sacrifice needed security solely for a larger test score.
If the VPN sometimes looks faster
A VPN-on test can occasionally exceed an off test because tests vary, routes differ or a provider treats particular paths differently. One such result is not proof that your ISP throttles traffic. Look for a consistent pattern using repeated comparable tests, consider the VPN's exit location, and avoid declaring traffic shaping without stronger evidence.
Judge by the task, not just the percentage
A 15% reduction might be unnoticeable on a high-bandwidth link, while a smaller loss may matter on a connection close to the bitrate needed for your stream. Games and video calls are sensitive to latency and jitter, not just Mbps. For streaming buffering, test the target application too; for ping and jitter, note the stability of each run.
Official perspective: Proton VPN's speed troubleshooting guidance covers server and protocol factors. NextWatch AI is independent and does not access or certify the VPN service you use.