What does this test measure?
A single speed number cannot explain why a video call freezes or an online game lags. This utility reports eight core metrics that collectively establish your true connection quality.
Download speed
How quickly data reaches your machine, expressed in megabits per second (Mbps). It dictates 4K streaming performance, rapid web browsing, application downloads, and software updates. Parallel HTTP connections are opened to fill high-bandwidth links and discount initial TCP slow-start ramps.
Upload speed
The speed at which your device pushes data out to the wider web. This is critical for zero-lag video conferencing, live streaming, cloud backups, and large file dispatching. Because most consumer fiber and cable links are asymmetric, the uplink is frequently an overlooked bottleneck.
Ping (latency)
The total round-trip time (RTT) for a packet to reach an edge node and return, measured in milliseconds. Sub-30 ms feels instantaneous, 30 to 60 ms is solid for competitive gaming, while latency climbing past 100 ms produces noticeable desynchronization.
Jitter
The statistical variance between successive ping samples. A connection showing a 25 ms average with 2 ms jitter delivers a reliable experience; that same 25 ms average with 45 ms jitter will drop voice packets and cause micro-stutters in real-time streams.
Packet loss
The percentage of transmitted data packets that never arrive at their destination. A loss rate of even 1% impairs voice communication and real-time gaming. On local networks, packet loss points directly to radio interference or low signal strength.
Bufferbloat
The unnecessary delay introduced by routers over-buffering packets during periods of heavy traffic. We benchmark latency while your download and upload streams are fully saturated. Spikes above 100 ms indicate that your router does not handle queue scheduling efficiently.
Consistency and call quality
Consistency indicates throughput stability over the duration of the transfer. The estimated Mean Opinion Score (MOS) calculates overall VoIP voice clarity on a 1.0 to 4.5 scale based on combined latency, packet loss, and jitter data.
WiFi score
A composite 0-to-100 quality rating computed from speed, baseline latency, jitter, packet loss, and bufferbloat metrics, ensuring that high-throughput connections with severe stability problems receive an accurate, realistic assessment.
The extra tests explained
Plan check
Compares real throughput against the bandwidth tier you pay your ISP for. On 5 GHz WiFi, achieving 60% to 85% of your plan's maximum speed is typical. Getting below 50% suggests testing over a physical Cat6 cable to isolate internal wireless degradation from external ISP issues.
DNS-over-HTTPS (DoH) benchmark
Queries encrypted DNS endpoints via Cloudflare, Google, Quad9, and OpenDNS. If these resolve in under 30 ms while standard browsing feels sluggish, your default ISP DNS server is likely overloaded or rate-limiting your queries.
LAN gateway responsiveness
Sends targeted sub-millisecond HTTP probes to common default local router addresses (such as 192.168.1.1 and 192.168.0.1). If your local gateway delay exceeds 15 ms, the source of your network slowdown is the local wireless connection rather than your broadband provider.
Payload scaling (MTU test)
Evaluates network behavior across increasing packet sizes (64B through 64KB). Sharp latency increases on larger payloads can signal packet fragmentation issues, mismatched Maximum Transmission Unit (MTU) settings, or aggressive TCP throttling along the route.
How to get an accurate result
- Test in your normal workspace: Run the diagnostic where you actually use your machine rather than standing right beside the router.
- Halt background data traffic: Pause torrents, cloud drive synchronizations, and streaming media across other devices on the network.
- Disable VPN tunnels: Running through an encrypted tunnel tests the VPN node's capacity and routing rather than your true physical link.
- Compare frequency bands: Run the test on 5 GHz or 6 GHz, then repeat on 2.4 GHz to determine the better band for your room.
- Conduct multiple runs: Take the median of three consecutive tests using the built-in history ledger.
- Verify against a wired cable: An Ethernet baseline immediately confirms whether poor throughput is caused by radio interference or ISP bandwidth limits.
How to read your results
Look at your metrics in context rather than focusing on a single number. If wired Ethernet tests reach full plan speed but WiFi throughput drops substantially, the bottleneck exists in the local wireless environment.
- High download speed with high jitter or packet loss: Indicates local radio interference or channel congestion. Switch channels or switch from 2.4 GHz to 5 GHz.
- Low speed paired with low latency: Suggests an ISP bandwidth throttle, plan limit, or peak-hour neighborhood node congestion.
- Fast speeds with severe bufferbloat: The router needs queue management (SQM) enabled to prevent line saturation from delaying latency-critical packets.
- Fast download with minimal upload: Normal for asymmetric cable or VDSL plans; symmetric speeds require a pure fiber link.
What is a good WiFi speed?
| Activity | Minimum | Recommended | Ping Target |
|---|---|---|---|
| Web browsing & email | 5 Mbps | 25 Mbps | <100 ms |
| 4K UHD streaming | 25 Mbps | 45 Mbps | <70 ms |
| Video conferencing (Zoom, Teams) | 5 Mbps up/down | 15 Mbps up/down | <50 ms, jitter <15 ms |
| Competitive online gaming | 10 Mbps | 30 Mbps | <35 ms, 0% packet loss |
| Cloud gaming (GeForce Now, Xbox) | 35 Mbps | 75 Mbps | <25 ms, jitter <5 ms |
| Multi-user remote office | 100 Mbps | 300+ Mbps | <40 ms |
WiFi standards and bands
| Standard | Industry Name | Frequency Bands | Max Theoretical | Typical Real-World |
|---|---|---|---|---|
| 802.11n | WiFi 4 | 2.4 / 5 GHz | 600 Mbps | 40 to 120 Mbps |
| 802.11ac | WiFi 5 | 5 GHz | 3.5 Gbps | 200 to 500 Mbps |
| 802.11ax | WiFi 6 | 2.4 / 5 GHz | 9.6 Gbps | 450 to 900 Mbps |
| 802.11ax | WiFi 6E | 2.4 / 5 / 6 GHz | 9.6 Gbps | 600 to 1,400 Mbps |
| 802.11be | WiFi 7 | 2.4 / 5 / 6 GHz | 46 Gbps | 1.5 to 3.5 Gbps |
How to fix slow WiFi
1. Optimize router placement
Place the router in an elevated, central location away from dense concrete walls, metallic surfaces, mirrors, and dense electronic equipment. Tucking a router inside a TV cabinet or on the floor degrades signal coverage.
2. Channel width and frequency optimization
Use 5 GHz or 6 GHz for work machines and gaming setups. Keep 2.4 GHz exclusively on non-overlapping channels (1, 6, or 11). If your 5 GHz band experiences frequent drops in a crowded building, reduce channel width from 160 MHz to 80 MHz to improve signal-to-noise ratios.
3. Deploy wired backhauls for mesh nodes
Wireless mesh repeaters that lack a dedicated backhaul radio can cut your available throughput by up to half on each wireless hop. Connecting access points back to the main router with Cat6 Ethernet or MoCA adapters preserves full line speed.
4. Resolve background interference
Common household appliances like microwave ovens, cordless phones, and unshielded USB 3.0 cables emit interference across the 2.4 GHz spectrum. Moving your router away from these sources helps eliminate sudden packet loss.
How to fix bufferbloat
Bufferbloat occurs when an internet link runs at maximum capacity, filling unmanaged router memory buffers and delaying interactive packets like gaming inputs or VoIP audio frames.
- Access your router's administration console via your gateway IP.
- Navigate to the QoS or Traffic Management panel and enable Smart Queue Management (SQM).
- Select algorithms such as CAKE or FQ-CoDel where available.
- Set your bandwidth limits to roughly 90% to 95% of your measured upload and download speeds, giving the router the margin needed to prioritize interactive packets.
How to check WiFi signal strength (dBm)
Web browsers cannot read raw radio hardware metrics directly. To check your true Received Signal Strength Indicator (RSSI) in dBm, use these system utilities:
- Windows: Open PowerShell or Command Prompt and execute:
netsh wlan show interfaces - macOS: Hold the Option key and click the WiFi menu bar icon to view RSSI and Tx Rate.
- Linux: Run terminal command:
nmcli dev wifioriwconfig
| Signal Level (dBm) | Connection Quality | Suitability |
|---|---|---|
| −30 to −50 dBm | Optimal | Maximum throughput, low latency |
| −50 to −65 dBm | Good | Smooth 4K video, reliable voice calls |
| −65 to −75 dBm | Fair | Basic web browsing, potential jitter |
| Below −75 dBm | Poor / Unstable | Frequent disconnects, severe packet loss |
How this test works and its limits
This test measures latency using precision high-resolution timestamps against edge server endpoints. Bandwidth is determined by streaming chunked binary arrays across concurrent HTTP/2 and HTTP/3 streams to fully utilize available network bandwidth. Upload tests dispatch random cryptographic binary blobs via parallel POST requests to test outbound capacity.
- Edge server proximity: Bandwidth is benchmarked against the nearest Cloudflare edge datacenter rather than every web destination you access.
- Browser sandbox overhead: High-performance multi-gigabit connections (above 1.5 Gbps) can encounter browser JavaScript engine bottlenecks. For multi-gigabit verification, native command-line tools like iPerf3 or native apps provide the most accurate reading.
Privacy and data handling
Test traffic is processed strictly in-memory between your browser and CDN edge endpoints. Diagnostic measurements (down, up, ping, jitter, bufferbloat) are saved purely inside your browser's local localStorage database and can be cleared at any time. We do not track, log, or resell your personal IP address.
Glossary
- Mbps vs MB/s:
- Mbps (megabits per second) measures network transfer speed; MB/s (megabytes per second) measures disk file size. Divide Mbps by 8 to calculate real-world file download speeds.
- Bufferbloat:
- Excessive latency that occurs when a high-volume download or upload fills up unmanaged router memory buffers.
- DoH (DNS-over-HTTPS):
- A protocol that encrypts standard DNS lookups inside HTTPS traffic to protect user privacy and prevent DNS tampering.
- SQM (Smart Queue Management):
- An intelligent router scheduling system (using algorithms like CAKE or FQ-CoDel) that balances packets to keep latency low under full network loads.
FAQ
Why is my speed lower than my plan?
Plans quote best-case wired speed. WiFi loses speed to distance, walls, interference, older devices, and shared airtime. Test over Ethernet to separate the two.
Does this test show WiFi signal strength?
No. Browsers block radio data such as dBm, channel, and band. This tool measures the result of your signal: speed, latency, jitter, and loss.
How much data does the test use?
Quick mode uses roughly 50 to 150 MB and Extreme can use 500 MB or more on fast links. Avoid metered mobile hotspots.
What is a good ping for gaming?
Under 40 ms with 0% loss and low jitter is ideal. Bufferbloat matters as much as raw ping.
Is 100 Mbps enough?
For most homes with a few users, yes. It handles 4K streaming, calls, and gaming together. Larger households or heavy uploads may want 300 Mbps or more.
Why do results change every time?
Neighbors, other devices, time of day, and server routing all change the result. Compare several tests.
Is 5 GHz always better than 2.4 GHz?
Faster, yes, but shorter range. In a far room 2.4 GHz can beat a weak 5 GHz signal.
Does a VPN affect the result?
Yes. A VPN adds distance and encryption overhead, so you test the VPN path. Turn it off for a true reading.
Why is upload so much lower than download?
Most consumer plans are asymmetric. Fibre plans are often symmetric.
Will a new router make my internet faster?
Only if the router is the bottleneck. It cannot exceed your plan speed, but it can improve WiFi range, stability, and bufferbloat.
Is it safe and private?
Yes. Test traffic goes to Cloudflare and a handful of public endpoints; your last 10 results stay in your own browser and can be cleared any time.