Which Ookla clone is the best?

TL;DR

Overall, we find the clones to perform just as well as the original Ookla Speedtest. We recommend using the Ookla clone https://speedtpe.speedtestcustom.com/.

This is because this clone is secure, uses a local server, doesn't show a logo, and has the best contrast ratio.

 

Contents

  1. Introduction
  2. Test approach
  3. Test results
 

Introduction

Even though we do not recommend testing with Ookla's Speedtest.net. There are plenty of reasons to use Ookla's Speedtest. The most important ones are:

  1. Ookla's speedtest is the gold standard
  2. Ookla's speed test is recommended by many ISPs
  3. Authorities recommend using Ookla's speedtest (and M-Lab)

If you really want to use Ookla's speedtest, you should use a clone instead of the original.

Accuracy, privacy, time to complete, data usage and the like are not important at this point. We do find the following criteria important:

  1. The clone is secure (https:// is used)
  2. Is it possible to select a nearby server
  3. Is the use of color such that the results are easy to read
  4. Is there a custom logo missing

The tacit assumption is of course that all clones measure as accurately as the original Ookla speed test.

 

Test approach

The tested Ookla clones are eligible for this study.

For each Ookla clone, we use the base URL (in most cases, you can find out what the base URL is with our Which speed test? tool).

Then, we investigate to what extent the selected speed tests meet the above criteria.

Finally, we test our internet speed based on the nearest server and compare the results with the original Ookla Speedtest.

 

Test results

The following Ookla clones allow to test you with a local server and match all other criteria.

  1. https://speedtpe.speedtestcustom.com/ (Color: #090F78, contrast ratio 15.54:1)
  2. https://tiscali.speedtestcustom.com/ (Color: #68149c, contrast ratio 9.68:1)
  3. https://fusionconnect.speedtestcustom.com/ (Color: #086bbc, contrast ratio 5.46:1)
  4. https://testatt.speedtestcustom.com/ (Color: #a04ad6, contrast ratio 4.72:1)

Note that these speed tests are sorted on the contrast ratio as calculated by WebAim's Contrast checker.

Accuracy

The selected server is VodafoneZiggo at the location Schiphol When we run Ookla's speedtest, the testserver "VodafoneZiggo Schiphol" is suggested. Hence we test the selected clones also against this testserver.

We measured the following speeds (with an advertised speed of 100 Mbps):

  1. Ookla: 101.46, 100.98, 101.32, 101.37 and 101.04 Mbps (average: 101.23 Mbps, standard deviation: 0.20 Mbps)
  2. Speedtpe: 102.0, 101.7, 102.0, 100.9 and 102.0 Mbps (average: 101.7 Mbps, standard deviation: 0.5 Mbps)
  3. Tiscali: 100.4, 101.1, 101.9, 101.9 and 101.9 Mbps (average: 101.4 Mbps, standard deviation: 0.4 Mbps)
  4. Fusionconnect: 102.0, 102.0, 102.0, 101.9 and 101.9 Mbps (average: 102.0 Mbps, standard deviation: 0.1 Mbps)
  5. Testatt: 102.0, 102.0, 102.0, 102.0 and 101.9 Mbps (average: 102.0 Mbps, standard deviation: 0.1 Mbps)
 

Findings

Based on the measurements above we conclude that:

  1. the ookla clones have one digit less than the original
  2. the clones measure a slightly higher speed than the original (on average 101.8 Mbps versus 101.2 Mbps)
  3. the standard deviation of the averages of the clones is 0.3 Mbps

Overall, we find the clones perform just as well as the original. We recommend using the Ookla clone with the most contrast (i.e. https://speedtpe.speedtestcustom.com/) as it is the easiest to read.