Run a browser-based packet loss estimate to check whether your connection is dropping or timing out requests. The tool samples repeated network requests, measures failed samples, latency, jitter, burst loss, and gives a connection quality grade.
Packet loss can cause lag spikes, rubber-banding, choppy voice, frozen video, cloud gaming stutter, and dropped calls even when your download speed looks fast.
Choose a sample size and start testing. The widget sends repeated same-origin HTTP requests with a timeout. Successful samples are shown as green/yellow bars based on latency, and failed or timed-out samples are shown as red bars.
Browser limitation: this is a failed HTTP request / timeout estimate. It is not raw ICMP ping, UDP packet capture, router telemetry, or an exact game-server packet loss test.
Packet loss happens when data traveling across a network does not reach its destination successfully. Every online game update, voice call syllable, video frame, website request, and cloud gaming input is broken into network packets or request data. When some of that data is lost, delayed too long, or discarded, real-time apps may stutter, glitch, or try to guess what was missing.
Packet loss is different from normal latency. Latency is delay. Packet loss is missing data. A connection can have low ping but still drop packets. That type of connection feels inconsistent: one moment the game is smooth, then suddenly movement snaps backward, voice cuts out, or a video call freezes. Loss can also happen in bursts, where several packets fail in a row. Burst loss is often more noticeable than isolated single-packet loss.
Packet loss can happen on your local Wi-Fi, inside your router, between your router and modem, on the ISP line, during congested routing, at a VPN endpoint, or near the remote server. It can also be caused by weak wireless signal, damaged cables, overloaded equipment, bufferbloat, network congestion, or unstable powerline adapters.
A fast download speed does not guarantee low packet loss. Speed tests often focus on throughput, while games and calls depend on reliability, latency, jitter, and packet delivery consistency. That is why a 500 Mbps connection can still feel bad for gaming if packets drop under load.
Use these ranges as practical guidance for gaming, video calls, VoIP, and cloud gaming. The lower the loss percentage, the better. Latency and jitter still matter, but packet loss is especially damaging because the application may never receive the missing data.
| Grade | Loss Estimate | Quality | Real-Time Impact |
|---|---|---|---|
| A+ | 0% | Excellent | Ideal for competitive gaming, calls, streaming, and cloud play. |
| A | 0.1–0.5% | Very good | Usually smooth; tiny isolated failures may not be noticeable. |
| B | 0.5–1% | Acceptable | May cause occasional stutter in sensitive games or calls. |
| C | 1–3% | Unstable | Noticeable rubber-banding, choppy audio, or video issues. |
| D | 3–5% | Poor | Gaming, VoIP, and cloud gaming may feel unreliable. |
| F | 5%+ | Severe | Frequent glitches, disconnects, and unusable real-time quality. |
This tool sends repeated HTTP requests from your browser to the current page with a unique cache-busting query string. Each request is timed with a browser timer and controlled with a timeout. If the request completes before the timeout, it is counted as successful. If it fails or times out, it is counted as a failed sample. The loss estimate is failed samples divided by total samples.
The test also records latency for successful samples. From those values, it calculates average latency, minimum latency, maximum latency, jitter, and slow samples. Jitter is estimated from differences between consecutive successful latency samples. Slow samples are successful requests that took much longer than normal, which may indicate congestion or unstable routing even when they did not fully fail.
This approach is useful for a quick web-based connection quality check, but it is not the same as a router-level UDP or ICMP packet loss test. Browser requests include website server response, CDN behavior, TLS/session handling, WordPress caching, browser scheduling, Wi-Fi conditions, and local device load. Your game may connect to a different server using a different protocol. Treat the result as a practical reliability estimate, then compare with game network graphs, router diagnostics, or ISP tools if problems persist.
A request completed before timeout and returned without a browser network error.
A request timed out, was aborted, or failed with a browser network error.
Failed samples divided by total samples, displayed as a percentage.
The longest streak of failed samples in a row. Burst loss often feels worse than isolated failures.
Your character moves forward, then snaps backward, or enemy positions appear to teleport.
Shots or abilities feel like they should connect, but the server never confirms them.
Discord, Zoom, Teams, or VoIP audio cuts out, drops syllables, or sounds robotic.
Video pauses, jumps, pixelates, or audio/video sync becomes unstable.
Video frames drop, controls feel inconsistent, or the stream quality suddenly falls.
Games, VPNs, remote desktops, or calls disconnect even though speed tests look normal.
Packet loss can be local, ISP-side, route-side, or server-side. The best troubleshooting path is to isolate where the loss begins. If loss appears only on Wi-Fi but not Ethernet, focus on wireless. If loss appears on Ethernet too, check router, modem, cables, ISP line, and congestion. If loss appears only in one game or service, the route or remote server may be the problem.
Start with simple isolation. Run the test over Wi-Fi, then run it again over Ethernet. If Ethernet is clean, your internet line may be fine and the problem is likely wireless. If both show loss, test with fewer devices, reboot equipment, check cables, and test at different times of day.
Wired testing removes most Wi-Fi interference and gives the cleanest baseline.
Power cycle equipment to clear temporary lockups, memory issues, or stale connections.
Try a known-good Ethernet cable and check that ports are not loose or damaged.
Pause cloud backup, torrents, livestreaming, and large uploads while gaming or calling.
Move closer, use 5/6 GHz where appropriate, reduce mesh hops, and change crowded channels.
Smart queue management can reduce congestion and bufferbloat under load.
Router firmware updates may improve stability, Wi-Fi performance, and packet handling.
If loss continues on wired tests at the modem/router, ask your ISP to check line quality and local congestion.
Latency, jitter, and packet loss are related, but they describe different connection problems. Latency is delay. Jitter is delay variation. Packet loss is missing data. Real-time apps need all three to be healthy: low enough latency, steady timing, and reliable delivery.
| Metric | Meaning | Measured In | What It Feels Like |
|---|---|---|---|
| Latency | How long data takes to travel. | Milliseconds | Input delay, delayed voice, slow response. |
| Jitter | How much latency changes between samples. | Milliseconds | Stutter, inconsistent ping, uneven calls. |
| Packet Loss | Data fails to arrive successfully. | Percentage | Rubber-banding, dropouts, missing audio, disconnects. |
| Burst Loss | Several failures occur in a row. | Consecutive lost samples | Sudden freeze, teleport, or call gap. |
Different apps tolerate loss differently. File downloads can retry missing data, so small loss may simply reduce speed. Real-time apps cannot always wait for retries, so loss creates visible or audible glitches. Competitive gaming, VoIP, video meetings, cloud gaming, remote desktop, and livestreaming are especially sensitive.
| Use Case | Ideal Loss | Warning Range | Problem Range | Impact |
|---|---|---|---|---|
| Competitive Gaming | 0% | 0.5–1% | 1%+ | Rubber-banding, poor hit registration, movement correction. |
| Cloud Gaming | 0% | 0.5–1% | 1%+ | Video artifacts, input inconsistency, stream quality drops. |
| VoIP / Voice Chat | 0–0.5% | 0.5–2% | 2%+ | Robotic voice, missing syllables, choppy calls. |
| Video Meetings | 0–1% | 1–3% | 3%+ | Frozen video, audio gaps, low resolution. |
| Downloads | Low | Varies | Severe loss | Retries, slower transfer, possible timeouts. |
Run multiple tests, not just one. A single test can be affected by a temporary website response delay, browser throttling, Wi-Fi interference, CPU load, VPN routing, or background traffic. A pattern across several tests is much more useful than one result.
Test in layers. First test while idle on Ethernet. Then test while idle on Wi-Fi. Then test while someone is streaming or uploading. Then test with VPN off and on. If packet loss only appears on Wi-Fi, focus on wireless. If it appears only under upload load, focus on bufferbloat, SQM, or QoS. If it appears all the time on Ethernet, check cables, modem, router, and ISP line quality.
For gaming, compare browser results with the game’s network stats. Many games show packet loss, ping, latency variation, or connection icons. If the browser test is clean but one game shows loss, the route to that game server may be the issue. If every service shows loss, the problem is likely local or ISP-side.
Wi-Fi can hide the real source of the problem. Ethernet gives a better baseline.
Packet loss or timeouts can appear only when uploads fill the connection.
High download speed does not guarantee stable packet delivery.
Different services use different routes. Test multiple tools when diagnosing serious issues.
VPN routing can add instability. Compare with VPN off.
One long burst of loss can feel worse than a few isolated lost samples.
This page covers packet loss test, connection quality checker, online packet loss test, packet loss checker, internet stability test, failed request estimate, gaming packet loss, VoIP packet loss, video call packet loss, cloud gaming packet loss, packet drop test, latency test, jitter test, burst packet loss, network reliability, Wi-Fi packet loss, Ethernet packet loss, bufferbloat, ping spikes, and connection diagnostics.
Packet loss happens when network data does not reach its destination successfully. In real-time apps, this can cause missing audio, rubber-banding, video freezes, and lag spikes.
No. Normal browser JavaScript cannot send raw ICMP ping packets. This tool estimates loss using failed or timed-out HTTP requests from the browser.
0% is ideal. Even around 1% can be noticeable in competitive games, especially if the loss happens in bursts.
Yes. Download speed measures throughput, while packet loss measures reliability. A fast connection can still drop packets due to Wi-Fi, congestion, routing, or hardware problems.
Wi-Fi can suffer from weak signal, interference, crowded channels, distance, walls, mesh hops, and competing devices. Ethernet is usually more stable for testing and gaming.
Use Ethernet, restart router and modem, replace cables, stop heavy uploads, improve Wi-Fi signal, enable SQM or QoS, update router firmware, test without VPN, and contact your ISP if wired loss persists.
Burst packet loss means several packets or requests fail in a row. It often feels worse than isolated single losses because apps lose a chunk of real-time data at once.
Yes. Lost or delayed voice packets can create choppy audio, robotic sound, dropped syllables, or gaps in conversation.
Results can change because of Wi-Fi conditions, server response, ISP congestion, background apps, router load, browser scheduling, VPN routing, and time of day.
No. Use this as a quick connection quality estimate, then compare with your game’s network stats or router diagnostics for the exact game-server route.
Run the packet loss estimate, compare wired and Wi-Fi results, and use the troubleshooting guide to reduce lag spikes, dropped calls, and unstable gameplay.