Measure your real download and upload speed with multi-connection accuracy, latency testing, and detailed history tracking.
Download Free on the App StoreSpeed test and all 19 tools are free. No ads, no account required.
Most speed test apps open a single TCP connection and report whatever throughput that one stream achieves. On connections faster than about 200 Mbps, a single connection often cannot fill the pipe due to TCP congestion control. PingKit opens six parallel connections simultaneously, saturating your link the same way a household full of devices would. The result is an accurate measurement of your actual available bandwidth, not an artificial bottleneck.
Six parallel TCP streams saturate even gigabit fiber and fast 5G connections. Get accurate results above 700 Mbps where single-connection tests fall short.
Separate download and upload measurements so you can see the full picture. Upload speed matters for video calls, cloud backups, and streaming.
Raw bandwidth is only half the story. PingKit measures connection latency alongside speed, showing whether your link is responsive enough for real-time tasks.
Every test is logged with timestamp and connection type. Compare WiFi vs cellular performance and track whether your ISP delivers consistent speeds over time.
Your ISP advertises "up to" a certain speed, but what are you actually getting? Run PingKit's speed test at different times of day to see if your provider delivers during peak hours. If you are consistently getting half the speed you pay for, you have data to support a call to your ISP or a switch to a competitor.
Sometimes your iPhone's cellular connection is faster than your WiFi. Run a speed test on both to find out. If WiFi is significantly slower, the bottleneck might be your router placement, channel congestion from neighbors, or an outdated router that cannot keep up with your internet plan.
Upgrading your router, changing DNS settings, or moving your access point? Run a speed test before you make changes and again after. PingKit's history makes it easy to compare results and confirm whether the change actually improved anything.
WiFi speed varies room to room depending on distance from the router, walls, and interference. Walk through your home or office running tests in each spot. You will quickly find the dead zones and the sweet spots, which tells you exactly where a mesh node or extender would help most.
For online gaming, latency matters more than raw download speed. A connection with 50 Mbps and 15 ms latency will feel snappier in a competitive match than a 500 Mbps link with 80 ms of lag. PingKit reports latency alongside throughput on every run, so you can tell whether jitter or distance to the server is the real culprit. If your speed looks fine but games still feel sluggish, our ping test and MTR tool dig deeper into where the delay is introduced along the route.
Zoom, FaceTime, Teams, and Google Meet lean heavily on your upload speed, which is often a fraction of your download speed on cable and DSL plans. If colleagues say you freeze or sound robotic, run a speed test and look at the upload number specifically. Anything below about 3 to 5 Mbps upload with high latency or jitter will cause trouble on group calls. PingKit measures upload as its own metric rather than burying it.
Streaming 4K HDR from Netflix, Apple TV, or YouTube needs a sustained 25 Mbps or more per stream, and several streams at once multiply that requirement. If video buffers or drops to lower resolution in the evening but tests fine in the morning, you are likely seeing peak-hour congestion. Running PingKit at different times builds a record you can use to confirm the pattern.
This is the single most common reason two speed test apps disagree on the same network. A single TCP connection is governed by TCP congestion control, which ramps throughput up gradually and backs off the moment it detects any packet loss or latency increase. On a slow connection that ceiling is rarely a problem, because one stream can fill the available bandwidth before congestion control intervenes. On a fast connection it becomes a hard limit: a single stream simply cannot accelerate quickly enough to use all of a gigabit fiber or fast 5G link before the test window closes.
The effect grows with speed and distance. The further the test server is from you, the higher the round-trip time, and the more a single TCP connection is throttled by its own window-scaling behaviour. That is why a one-connection test might report 180 Mbps on a line that is genuinely delivering 900 Mbps. The line is fine; the measurement method is the bottleneck.
PingKit opens six parallel connections and adds their throughput together. This mirrors how your network behaves in real life, where dozens of apps and devices share the link at once, and it is the same approach professional and ISP-grade testing services use. We picked six because it reliably saturates consumer and prosumer links without overstating results through excessive parallelism. The goal is an honest measurement of what your connection can actually deliver, not an inflated headline number.
Rule of thumb: if PingKit reports a higher number than another app on the same network, it is usually because the other app used a single connection and could not fill your link. If PingKit reports a lower number, check whether the other app quietly selected a much closer server, or whether something on your network was using bandwidth during the test.
There are plenty of free speed test apps in the App Store. Most do one thing and surround it with advertising. Here is an honest comparison of where PingKit differs from the typical free speed-test app.
| Capability | PingKit | Typical free speed-test app |
|---|---|---|
| Multi-connection accuracy | Six parallel connections | Often a single connection |
| Ads | None | Banner and interstitial ads common |
| Account required | No account, no sign-up | Often required for history |
| Bundled tools | 19 network tools in one app | Speed test only |
| Scheduled / automatic tests | Yes, via Guardian + free Mac Agent | Manual runs only |
We are not claiming PingKit is the only good speed test on iPhone. The point is that a speed test on its own rarely answers the question you actually have, which is usually "why is my connection slow?" Bundling the speed test with traceroute, ping, MTR, DNS lookup, port scanning, and continuous monitoring means you can go from symptom to cause without juggling five separate apps, each with its own ads and account.
First, rule out the measurement method. If you are comparing against a single-connection app, PingKit's multi-connection result is the more accurate one. If PingKit itself shows well below your plan, check the basics: are other devices streaming or backing up to the cloud, is your iPhone on the 2.4 GHz band when 5 GHz is available, and how far are you from the router? Wire a computer directly to the modem if you can and compare. A large gap between wired and WiFi results points to a WiFi problem; a large gap on a wired connection points to the ISP. Our guide on how to tell if your ISP is throttling walks through the next steps.
Some variation is normal. Shared media like cable and 5G fluctuate with how many neighbours are online, and your iPhone's radio conditions change as you move. To get a representative figure, run three or four tests in a row and look at the range rather than a single number. PingKit's history makes this easy because every run is timestamped. Wildly inconsistent results, especially with spikes in latency, often indicate interference or a flaky connection rather than a true bandwidth problem, which is where our connection monitor earns its keep by watching the link continuously instead of for a few seconds.
If WiFi tests slower than cellular on the same phone in the same spot, your internet plan is probably not the limiting factor; your WiFi is. Common causes are an old router that cannot handle your plan's speed, channel congestion in dense buildings, and physical distance or walls between you and the access point. Test in the same room as the router to confirm, then test from where you actually use the phone to measure the real-world drop-off.
Distance to the test server adds latency and can lower the result on a single-connection test in particular. When you want to compare apps fairly, make sure they are both testing against a server in roughly the same location. A test to a server on another continent will always look slower than one across town, and that difference reflects geography rather than the health of your connection.
A speed number alone does not tell you much. PingKit combines speed testing with 18 other network diagnostic tools. If your speed test shows unexpectedly low results, jump into Connection Monitor for continuous stability tracking, run a Traceroute to see where packets are being slowed, or use Smart Diagnostics for an automated analysis of what is wrong and how to fix it. Everything works together so you get answers, not just numbers.
Manual tests are great for spot checks, but they only capture the moment you happen to run them. The slow-down you want to catch usually happens while you are asleep or away from your phone. PingKit Guardian ($2.99/mo or $24.99/yr) adds scheduled, automatic speed tests that run on a recurring basis through the free PingKit Mac Agent, building a continuous record of your connection without you lifting a finger. Guardian also layers in AI-powered explanations of your results and 24/7 Mac monitoring that alerts you when your connection degrades. The speed test itself stays completely free; Guardian is for people who want it running on autopilot in the background.
PingKit uses six parallel TCP connections to saturate your link, the same technique professional speed-testing services use. A single connection often cannot fill a fast pipe, which is why many apps under-report on connections above 200 Mbps. The multi-connection approach gives accurate results even on gigabit fiber and fast 5G links exceeding 700 Mbps.
Most speed test apps use a single TCP connection, which is limited by congestion control and cannot fully utilize fast links. PingKit opens multiple parallel connections to measure your true maximum throughput. If PingKit shows higher speeds, it likely reflects your actual bandwidth more accurately. If it shows lower speeds, check whether the other app selected a closer server or whether something on your network was using bandwidth during the test.
Yes. Run a speed test on WiFi, then switch to cellular and run it again. PingKit logs every result with a timestamp and connection type, so you can compare WiFi and cellular performance side by side in your speed history.
For general browsing and social media, 25 Mbps is plenty. HD video streaming needs 10 to 25 Mbps, while 4K streaming wants 25 Mbps or more per stream. Video calls work best with at least 3 to 5 Mbps upload. Online gaming depends more on low latency than raw speed. PingKit tests all three metrics, so you can judge whether your connection fits your needs.
Run three or four tests in a row and look at the range rather than any single result. Shared connections like cable and 5G fluctuate naturally, so a small spread is normal. A large, erratic spread, especially with latency spikes, usually points to interference or a flaky link rather than a true bandwidth limit.
A speed test transfers data to measure throughput, so on cellular it will consume some of your data allowance, more on faster connections. This is true of every speed test app. If you are on a metered plan, prefer testing over WiFi, and avoid running back-to-back tests unnecessarily.
Asymmetric speeds are normal on most cable and DSL plans, which are designed to prioritize download. Fiber connections are often symmetric. Low upload only becomes a problem for video calls, cloud backups, and live streaming. PingKit reports upload as its own metric so you can see it clearly rather than having it averaged into a single figure.
Yes, with Guardian. The free version runs tests on demand. Guardian adds scheduled, automatic speed tests through the free PingKit Mac Agent, so your connection is checked on a recurring basis and logged to history without manual input.
Yes. The speed test and all 19 network tools in PingKit are completely free, with no ads and no account required. The optional Guardian subscription adds AI-powered analysis, 24/7 Mac monitoring, and scheduled speed tests, but the speed test itself is fully functional without it.
Continuous TLS cert monitoring, ISO 27001 PDF export, Network History AI, Apple Watch complication, and unlimited AI. $4.99/mo.
Learn about Guardian Plus →Download PingKit free and find out what your connection actually delivers.
Download Free on the App StoreRequires iOS 17.0 or later.