Quick test transfers up to 10.5 MB of payload in total, plus network overhead. Start only if you agree to use this data.
No location permission or installation needed. Test traffic goes to an external measurement server, which can see your IP address. This tool does not request your personal files or store your results.
Close other downloads and keep this page visible for a cleaner measurement.
Ready. Nothing is measured until you choose Start.
Mbps
Waiting for a real transfer
Dial maximum: 100 Mbps
Your connection snapshot
Each summary needs at least two trials of 50 ms or more. Repeated largest payloads take priority. Short samples produce a limited estimate; Standard or Extended can improve confidence.
Browser network timing is preferred; download server processing is excluded when available. Upload includes server acknowledgement. Single-request estimates, not a certified line-capacity test or ICMP ping.
Transfer trial comparison
Download · blue / Upload · purpleEach bar is one completed trial, in test order. Short trials appear with lower opacity and do not count towards the summary.
View all completed measurements
| Trial | Direction | Payload | Measured duration | Mbps | Summary eligible | Timing method |
|---|
No HTTP latency samples yet.
File transfer time estimator
Download: --
Upload: --
Uses this test's measured summary speed. Decimal units: 1 GB = 1,000 MB. Actual transfers may take longer because speed, server load and connection conditions change.
Internet Speed Test Online: Check Download, Upload and Response Time
Run a browser-based internet speed test to see how quickly your connection transfers data at this moment. Download speed describes incoming data, while upload speed describes outgoing data. HTTP latency and timing variation add context about responsiveness to the measurement server. Together, these readings offer a useful snapshot rather than a guarantee of performance on every website or app.
Get a cleaner speed reading
Pause large downloads, cloud backups and video streams before starting. Keep this page in the foreground, and avoid running another speed test at the same time. For a useful comparison, test the same device in the same location, then change only one factor, such as Wi-Fi position or a wired connection.
Mbps is not the same as MB/s
Mbps means megabits per second. MB/s means megabytes per second. Since one byte contains eight bits, divide a speed in Mbps by eight to get the equivalent MB/s. For example, 80 Mbps is 10 MB/s before other transfer overhead. Use the unit selector to view either format without restarting your test.
Understand response timing
The latency reading is the median time for a small HTTP request and its response, after a warm-up request. Timing variation is the average absolute difference between consecutive successful samples. Browser scheduling, server response and your network all influence these values. They are not a direct measurement of a game's ping or voice-call packet jitter.
Choose depth, not just a bigger number
Quick mode uses smaller transfers and is the default for lower data usage. Standard and Extended modes use larger samples that can provide a more useful estimate on faster connections. The result uses the median of up to three largest eligible trials, not the highest spike. A test may finish a direction early after three sufficiently long trials.
Internet Speed Test Questions
Why is my speed lower than my broadband plan?
Your plan may advertise a line speed measured differently from this browser test. Wi-Fi quality, shared traffic, distance to the measurement server, VPN routing, browser overhead and single-request transfer limits can all lower a result. Repeat under comparable conditions and compare with your provider's own diagnostic method before drawing conclusions.
Why does a result show two dashes?
A speed summary requires at least two successful trials lasting 50 ms or more in that direction. Extremely short trials are excluded because browser timing error can dominate them. Short but eligible samples are labelled as a limited estimate. Try Standard or Extended if your data allowance permits. The live dial may show a provisional trial speed before enough summary samples are available.
Does this test consume mobile data?
Yes. It downloads test bytes and uploads generated test bytes. The selected mode displays its maximum planned payload before you start. Network overhead adds to that amount, and stopping a request cannot recover bytes already transferred. Upload attempted counts requested payload, not confirmed bytes on the wire for a cancelled request.
Can it measure packet loss or loaded latency?
No. Failed HTTP requests are not a packet-loss percentage, and this tool does not measure timing while deliberately saturating the connection. Its timing samples are taken before the transfer trials. Download and upload throughput alone cannot determine gaming or call quality.
What happens if I switch tabs or lose the connection?
The tool stops when the page becomes hidden or the browser reports going offline. Completed samples remain visible and are marked as partial. Start again for a new measurement. If a direction times out or is blocked, successful results from other stages can still be shown.
Why can speed tests disagree?
Different tools use different servers, sample sizes, timing methods and numbers of simultaneous connections. This tool measures effective HTTP throughput over sequential single requests. It cannot certify your total line capacity. The underlying transfer endpoints are documented in the measurement project's reference.
Are my files or test results saved?
Uploads contain generated random test bytes, not files from your device. This tool keeps completed measurements only in page memory unless you copy or download them yourself. Clear results removes them from the tool. The external server still receives the test requests and your network address as part of normal internet traffic.