Seven scenarios. Specific numbers.
This table shows published upstream project results dated 29 October 2026. They compare its engine against Internet Download Manager 6.42. The LOD edition uses that code, but these figures are not a separate new benchmark of our Windows package.
| Scenario | IDM · s | Upstream engine · s | Ratio |
|---|---|---|---|
| 1 GiB · fast local link | 5.456 | 1.009 | 5.41× |
| 100 MiB · limited throughput | 10.234 | 6.319 | 1.62× |
| 100 MiB · 150 ms latency | 7.622 | 3.844 | 1.98× |
| 32 MiB · one slow connection | 7.213 | 3.432 | 2.10× |
| 50 MiB · two-connection limit | 15.864 | 12.568 | 1.26× |
| 50 MiB · interrupted connections | 5.428 | 1.647 | 3.30× |
| 256 KiB · small file | 3.141 | 0.030 | 105.90× |
Values are median total completion times. Lower is better. The ratio compares IDM time to upstream engine time; the report derives it from measurements, so it can differ slightly from dividing the rounded medians.
How the test was run
- Date: 29.10.2026.
- CPU: Intel Core i5-13400F; 63.8 GB RAM.
- System: Windows 11, build 26200.
- IDM: 6.42.25.3.
- Upstream revision:
caa200c. - Server: local labserver at 127.0.0.1:8097 serving generated content.
- Settings: both clients use their default settings, initially eight connections.
- Measurement: seven runs, median and interquartile range. The report includes confidence intervals and Mann–Whitney p-values.
What these numbers do not promise
The local test measures client behaviour under controlled conditions, including startup, transfer and completion time. It does not measure the maximum speed of an arbitrary internet website.
The small-file advantage is heavily influenced by startup delay. We do not present it as universally “106 times faster internet”. A slow server, account restriction, busy disk or network cap can dominate the result.
Inspect the data yourself
The files are included on this website without redirecting to the original project. The published test originates with the upstream project, as recorded in the attribution notices.
Back to the interactive comparison →