How to use this tool
- Choose the call type you expect to use. Simultaneous calls means separate calls sharing this internet connection, rather than participants inside one meeting.
- Pause downloads and backups, read the data-use note and press Start Measurement. No camera or microphone access is requested.
- Compare completed upload and download throughput with the selected reference. Review idle latency and the change under download or upload load as separate context.
- For a real problem, join the meeting application’s test call or inspect its own call statistics. Repeat here after changing one network condition.
A useful check before a meeting
A video call sends your camera and audio upstream while receiving other participants downstream. An internet plan’s advertised download speed says little about its upload capacity. This page measures both directions with the same HTTP capacity core as our homepage and compares the final numbers with published service guidance.
The comparison is intentionally narrow. It does not open Zoom or Meet, request media-device permissions, simulate the application’s codec or measure WebRTC packet loss. A route to Cloudflare is not the same route as a meeting provider’s media server.
Choose the correct published reference
Zoom’s Web App guidance distinguishes a 1:1 720p call from a 720p group call: 1.2 Mbps in each direction for the former, versus 2.6 Mbps upload and 1.8 Mbps download for the latter. Its listed 1080p reference is 3.8 Mbps upload and 3.0 Mbps download.
Google Workspace lists up to 1.7 Mbps in each direction for an individual’s 720p video, and up to 3.6 Mbps in each direction for 1080p. The options above identify those exact modes. These are reference values rather than promises that a particular account, device or meeting will send that resolution. Both services can adapt their bandwidth use.
Capacity headroom is only one part of call quality
Measured headroom is this run’s throughput minus the selected per-call reference multiplied by your chosen number of simultaneous calls. A positive number means the measured capacity was above that bandwidth reference at this moment. It is not a reserved channel for your call, and other household traffic can use that capacity.
Latency under load adds another view: a cloud backup or upload may make requests take longer even though measured Mbps exceeds a service’s reference. Review both download-loaded and upload-loaded delay. We show those measured differences without converting them into a universal call-quality score.
When a call still freezes
Check the meeting application’s own connection statistics and make a short test call from the actual room and device. CPU load, an overloaded browser, wireless interference, service routing and packet delivery can all affect a meeting without being diagnosed by this page.
Repeat using Ethernet if practical, pause a competing upload, or try a lower video resolution and compare. Use the bufferbloat comparison to focus on added delay during traffic, or the Wi-Fi room comparison to record measurements around your home. For planning several users, try the household bandwidth calculator.
Frequently asked questions
Does this certify my connection for Zoom, Meet or Teams?
No. It compares a measured HTTP capacity snapshot with selected published Zoom and Meet bandwidth references. It does not certify a meeting service, measure its actual media route or evaluate Teams-specific requirements.
Do I need to allow my camera or microphone?
No. This tool transfers measurement payloads and times HTTP requests. It does not request camera, microphone, location or conferencing-account access.
Why can I have enough Mbps but a poor call?
Available bandwidth is only one factor. Delay, variation, lost media packets, Wi-Fi contention, device performance and the meeting service can still affect quality. Check the application’s own call information alongside this test.
Primary sources and further reading
Explore all connection tools · Read our test methodology · Run the main speed test