Quick answer
Internet speed test results explained in plain English
Download speed tells you how quickly content reaches your device. Upload speed tells you how quickly you can send data back out. Ping measures delay. Jitter measures whether that delay changes from moment to moment. Fast download with bad ping can still feel sluggish, and low ping with weak upload can still ruin video calls.
Mbps and MB/s are different units. Internet plans and this test use megabits per second (Mbps). File downloads may show megabytes per second (MB/s): divide Mbps by eight. A 100 Mbps connection has a theoretical ceiling of 12.5 MB/s before network overhead; a slower file download alone does not prove your plan is underperforming.
Simple rule
- Download: matters most for streaming, browsing, and large file downloads.
- Upload: matters most for Zoom, Teams, cloud backups, and sending large files.
- Ping: matters most for gaming, calls, and anything interactive.
- Jitter: matters most when a connection feels inconsistent rather than simply slow.
Measure your connection, then read the results
Use the device and room where you notice the problem. Pause large downloads and backups, then compare two or three quiet tests. No account or app required. Opening the tool does not start a measurement.
Prefer the standalone tool? Open the full speed-test page.
Benchmarks
Judge each metric against a task and a comparison
One universal strong/usable/bad table hides the destination, workload, device, and service settings. Use this decision table to choose the right comparison, then consult the versioned activity requirements for planning values.
| Metric | Question it answers | Best comparison | Do not infer |
|---|---|---|---|
| Download | How much data this path receives per second | Sum simultaneous activity requirements; compare wired and Wi-Fi with the same endpoint | That a high score guarantees low lag, Wi-Fi coverage, or every service route |
| Upload | How much data this path sends per second | Current call, broadcast, backup, and concurrent-upstream requirements | That download capacity describes outgoing video or file completion time |
| Idle latency | How quickly the selected probe responds before heavy traffic | Repeated samples to the same endpoint plus the real application's telemetry | That the game, call region, or remote host uses the same route |
| Loaded latency | Whether responsiveness degrades during upload or download | Quiet and busy runs with the same device, endpoint, and household load | That one busy sample proves ISP congestion or local bufferbloat |
| Jitter | How much latency varies across samples | The median, spread, and in-app stability graph over several trials | That one sample describes long-term stability |
| Application probes | Whether the browser's timed probe requests completed | Repeated runs plus device, Wi-Fi, provider, and application evidence | Physical-line packet loss without a lower-layer measurement |
Need task-specific targets? Compare your result against the versioned activity requirements and evidence table, or inspect the immutable JSON release.
Diagnosis
How to interpret common result patterns
High download, high upload, low ping
Capacity and idle responsiveness look healthy to this endpoint. Compare with the simultaneous activity total, loaded latency, and each real application before calling the whole connection healthy.
High download, low upload
Received traffic may have headroom while outgoing calls, backups, or files compete for a smaller upstream path. Compare upload with the actual concurrent workload and provider label.
Good Mbps, bad ping or jitter
Capacity is not the immediate constraint in this sample. Compare Ethernet with Wi-Fi and quiet with loaded runs to separate radio conditions, queues, routing, and endpoint load.
Low download, low upload, low ping
This path appears responsive while its transfer capacity is limited. Compare with the activity table and provider label; the result alone does not identify the access technology or cause.
Plan vs actual speed
Why your real result may be lower than the advertised plan
For a US home internet plan, open your provider's Broadband Consumer Label and compare both typical download and upload speeds. A plan's headline download number does not describe upload performance or Wi-Fi in every room.
- Wi-Fi adds interference, range limits, and device-to-router variation that Ethernet avoids.
- Your household shares bandwidth. Active streams, updates, or console downloads can skew a test instantly.
- Routers, modems, and even laptop network cards can cap the top speed before the ISP does.
- Neighborhood congestion and routing conditions can raise ping or reduce throughput at busy hours.
Best way to validate your ISP speed
- 1. Plug a capable laptop or desktop into Ethernet.
- 2. Pause major downloads and cloud syncs.
- 3. Run several tests and keep the median and full range.
- 4. Test again on Wi-Fi in the rooms you actually use to spot in-home issues separately from ISP delivery.
Next actions
Use the result to decide what to fix next
A speed test becomes useful only when it leads to the right next step. If your biggest problem is buffering, focus on download and Wi-Fi coverage. If work calls are bad, focus on upload and jitter. If games feel delayed, focus on ping first and Mbps second.