Download speed
How quickly this device receives data. It matters most for streaming, downloads, updates, and loading media. Reported in Mbps.
Free connection diagnostic
Internet Speed Test
Measure download, upload, ping, jitter, and responsiveness under load. Compare the result with your plan and see what to test next.
Starts only when you chooseData Saver optionLocal result history
Cloudflare receives the test traffic and sees your public IP. Standard schedules up to 131.4 MB; overhead or retries can add data. Results and history stay in this browser.
Add test contextNot specified · Plan not entered
Internet—speed test
Test status
Ready when you are
- Download
- —Mbps
- Upload
- —Mbps
- Ping
- —ms
- Jitter
- —ms
For a clean baseline, pause downloads and connect with Ethernet or sit near your router. One click measures speed and responsiveness.
More than one big number
What this speed test actually measures
Throughput tells you how much data can move. Latency tells you how quickly the connection reacts. You need both stories to understand streaming, uploads, calls, games, and everyday browsing.
Upload speed
How quickly this device sends data. Video calls, cloud backups, file sharing, and live streaming depend on it.
Ping and jitter
The browser estimates HTTP round-trip response time, shown here as ping. Jitter is how much that timing varies. Lower, steadier values feel more immediate.
Loaded latency
Response time while the line is busy. A large increase can reveal queueing delay that a quiet ping and high Mbps miss.
The path your result includes
browser → test edgeBrowser & device
Wi-Fi or Ethernet
Router & modem
ISP & route
Cloudflare edge
This is an end-to-end browser measurement, not a direct reading from your modem. That makes it realistic for this device—but one run alone cannot prove which link caused a slow result.
Useful speed, not maximum speed
How much internet speed do you need?
Published minimums describe one activity. A practical target needs headroom for every other person, device, sync, and update using the connection at the same time.
ActivityStarting pointWhat the number misses
One 4K video stream
15+ Mbps downNetflix’s published recommendation. Add headroom for other streams and active devices.
HD video calls
3–5 Mbps both waysStable upload, sub-100 ms latency, low jitter, and low loaded latency matter as much as spare download capacity.
Online gaming
10/3 Mbps + low delayA Blockify diagnostic starting point, not a publisher requirement. Ping, jitter, loaded latency, server distance, and packet loss shape play more directly.
Busy connected household
100/20 Mbps baselineThis matches the FCC’s advanced fixed-broadband benchmark; it is a policy baseline, not a universal personal requirement.
Sources and context: Netflix speed recommendations, Google Meet network guidance, and the FCC 100/20 Mbps benchmark. Blockify’s “Ready” labels intentionally add headroom and are practical guidance, not service certification.
Read the pattern
What to test next when something looks wrong
These are diagnostic clues, not automatic verdicts. Change one thing at a time, repeat the test, and let the difference point you toward the next check.
Wi-Fi is slow, Ethernet is fast
Your ISP capacity may be fine. Test beside the router. If both the router and device support them, compare 5 or 6 GHz nearby; they can be faster but generally lose more reach through walls than 2.4 GHz. Reposition a mesh node and check device limits or interference.
Both Wi-Fi and Ethernet are slow
Pause other traffic, restart the gateway, test another device, and repeat off-peak. Persistent gaps are better evidence for an ISP conversation.
Fast download, weak upload
Downloads may feel fine while calls, backups, and live video struggle. Check whether another device is syncing and compare with your plan’s upload rate.
Fast Mbps, poor loaded latency
The line may become unresponsive when busy. Look for router SQM or smart queue management, stop saturation, and compare upload-loaded versus download-loaded delay.
A three-run evidence workflow
Turn a snapshot into useful evidence
Blockify keeps up to five results on this device. When the latest three use the same setup, mode, and method version, it shows their median download speed, range, and spread. Mixed setups remain a controlled comparison rather than being blended into a misleading average, and you can export every saved measurement for deeper review.
- RUN 01Establish the clean baseline
Use Ethernet if possible—or sit beside the router. Pause downloads, backups, streams, and VPNs you can safely stop.
- RUN 02Change one variable
Keep the same device and test mode, then move to the slow room or change from Ethernet to Wi-Fi. Label the setup in Test settings.
- RUN 03Check whether the pattern repeats
Repeat the setup that matters at the problem time. Review Recent tests for the three-run median and range, or export the CSV for jitter and loaded-latency details, before drawing a conclusion.
A different kind of speed
Fast connection, but websites still feel heavy?
Connection speed and page speed are different. This test cannot prove that an extension increases the Mbps sold by your ISP. Blockify can prevent some ad and tracker requests from loading on supported pages, reducing unnecessary page activity and making browsing feel cleaner.
Methodology, privacy & limits
How Blockify measures and interprets a run
Blockify maintains the custom sequence configuration and the interpretation labels documented below; they are not Cloudflare ratings. Methodology version 1.1 was updated on July 27, 2026.
Custom measurement sequence
After you click Start, Blockify uses the open-source Cloudflare engine for a custom HTTPS sequence to Cloudflare's edge. Standard schedules 12 idle-latency probes, 10 download samples, and 8 upload samples from 100 KB to 25 MB each (131.4 MB of nominal payload). Data Saver schedules 8 latency probes, 8 download samples, and 8 upload samples from 100 KB to 5 MB each (25.4 MB of nominal payload). In both modes, the first 100 KB sample in each direction is a calibration round that cannot end that direction early.
Responsiveness bands
An overall grade requires idle, download-loaded, and upload-loaded latency; missing values are shown as partial or not measured. With all three present, Blockify takes the higher loaded value, subtracts idle ping, and never reports a negative increase. The label is Excellent at 15 ms or less, Good above 15 through 40 ms, Sensitive above 40 through 100 ms, and Congested above 100 ms.
Plan comparison bands
The optional plan check divides measured download speed by the download speed you enter, then rounds the percentage: Above at 110% or more, Near at 90% through 109%, Worth a wired retest at 70% through 89%, and Well below under 70%. It is a diagnostic comparison, not a service-compliance decision.
Activity readiness bands
4K streaming is Workable at 15 Mbps download and Ready at 25 Mbps. A busy household is Workable at 50/10 Mbps and Ready at 100/20 Mbps. Video calls are Workable at 2/2 Mbps or more, with ping no more than 200 ms, jitter no more than 40 ms, and loaded increase no more than 100 ms. Ready requires 5/5 Mbps or more, with limits of 100 ms, 20 ms, and 40 ms respectively. Gaming is Workable at 5/1 Mbps or more, with limits of 100 ms, 40 ms, and 100 ms; Ready requires 10/3 Mbps or more, with limits of 50 ms, 20 ms, and 40 ms. Every listed condition must pass. Missing required values produce Not measured; measured results below Workable are Limited.
What stays local
Blockify stores recent measurements only in local browser storage. Page analytics and the engine’s completed-results logging are disabled on this route. Export happens only when you choose it.
Limits and ownership
One browser run cannot isolate a faulty cable, Wi-Fi interference, long-term uptime, or every route. Blockify does not present this custom sequence as a calibrated lab instrument or ISP line certification. Packet loss is not shown because this page does not run a supported UDP loss test.
Measurement engine: Cloudflare Speedtest (MIT-licensed). Cloudflare supplies the engine and routed edge; Blockify configures the limited sequence, interpretation bands, activity labels, and methodology version. The test starts only after your click, and final-result logging is off.
Privacy note: test requests necessarily expose your public IP to Cloudflare just as other internet requests do. Read Cloudflare’s privacy policy and Blockify’s privacy policy.
Related network checks
Get the rest of the connection picture
Speed test FAQ
Common internet speed questions
The test measures download throughput, upload throughput, idle latency (ping), jitter, and latency while download or upload traffic is active. Together, those measurements describe both how much data your connection can move and how responsive it stays while busy.
Good depends on what is happening at the same time. Netflix recommends at least 15 Mbps for one 4K stream, while the FCC uses 100 Mbps download and 20 Mbps upload as its current benchmark for advanced fixed broadband in the United States. A household with several active devices needs more headroom than one person watching a single stream.
It measures the complete path from this browser and device, through Wi-Fi or Ethernet, your router and ISP, to Cloudflare’s edge. A low result can come from any part of that path. Compare an Ethernet or near-router baseline with a test where the problem happens to narrow it down.
Tests can use different servers, routes, connection counts, file sizes, and calculation methods. Wi-Fi signal, other household traffic, device limits, VPNs, and time-of-day congestion also change. Compare repeated tests made with the same setup instead of treating one number as permanent.
Raw throughput is only one part of performance. High loaded latency, jitter, Wi-Fi interference, packet loss, a distant service, or a busy device can make calls, games, and pages feel delayed even when download speed is high. This test reports loaded latency so that the hidden responsiveness problem is easier to spot.
A successful run schedules no more than approximately 132 MB of test payload in Standard mode or 26 MB in Data Saver mode, and may use less. Protocol overhead or automatic retries can add data, so these are estimates rather than hard billing caps. Data Saver can under-read very fast connections, especially on higher-latency satellite, mobile, or VPN paths, because it transfers smaller samples. Avoid either mode on a metered connection unless you are comfortable with that usage.
The completed measurements and recent-test history stay in your browser unless you copy or export them. Blockify disables page analytics on this route and disables the measurement engine’s final-results logging. Test requests still go to Cloudflare, whose server can see the public IP and network information normally exposed by an internet request.
Yes. A VPN adds an encrypted route through another server, which can change both speed and latency. If you are diagnosing your ISP connection, compare a VPN-off baseline with a separate VPN-on run and label each setup. Do not turn off a required work or security VPN without permission.
Pause large transfers, use Ethernet or sit near the router, and take at least three readings with the same setup. Then repeat where the problem occurs or during the time it usually appears. The pattern across controlled tests is more useful than the best single score.
Contact your provider when repeated Ethernet tests stay well below the expected plan speed across different times and devices after you restart the modem and router. Export the local test history, note the setup and times, and ask the provider to test the line or gateway—not just Wi-Fi coverage.