The Best Free Antivirus for Windows – January 2016

I have been looking into the question of which free antivirus is the best for almost 10 years now, and the only way to answer is by looking at comparative tests with hard data, keeping in mind that the programs themselves are constantly changing and being updated, and that the picture that emerged from last year’s tests, say, may have changed with the more up-to-date round of tests, etc.

We will turn to “av-comparatives.org” for our data, since they regularly run what are arguably the web’s best and most comprehensive antivirus tests. As I write this in Jan 2016 I will use the latest versions of their tests, conducted within the last 6 months of this writing.

The contenders: a scan of several of the tests they conducted reveals a clear picture of who the ‘real’ players are. For free antivirus programs, the top three are Avira, AVG, and Avast. (Note: I debated whether to include Panda free antivirus, which seems like it wants to be a contender, but decided to skip it as it is clearly ‘not there yet’). In order to provide a frame of reference, I will also include the test results of the top three paid antivirus programs, which can also be identified with a cursory inspection of the results: Kaspersky, BitDefender, and Eset. I will also copy the practice from av-comparatives, of considering Microsoft Security Essentials as a control, the idea being that since it is now provided pre-intalled with on many Windows PC’s, you would need to install another antivirus program in it’s place only if it is the case that this new program provides significant additional protection.

The question I am looking answer:  simply, what is the best free antivirus program, and how does it fare compared to the paid programs.

The evaluation criteria: generally speaking, the three things I cared about were (1) The detection rate: i.e. how likely it is to catch viruses, (2) Performance: i.e. how much of the system resources did it drain; and (3) The rate of false positives: i.e. how likely it is to ‘catch’ things that it shouldn’t. All of this info was available in the av-comparatives data, although the tests didn’t necessarily correspond one-to-one with my criteria and other variables were also involved.

The tests that we looked at: we as a rule, and because we were writing this in January 2016, we only wanted to look at tests conducted within the last 6 months of 2015 (there were no tests conducted in 2106).

Also, because this is an investigation into the best free antivirus, we excluded any test that did not include all three free antivirus programs (Avira, AVG, and Avast).

Therefore, we looked at the following tests. Only the ones below fit our criteria

  1. Real-World Protection Test August – November 2015
  2. Performance – Test October 2015
  3. File Detection and False Positives Test – September 2015

Below, we will look at each one of these tests in turn. Note that we are using a color coding system in the tables as follows:  Green=Hight/Best ,  Orange=Middle , and   Red=Low .

1. Real World Protection Test (AUG-NOV 2015)

In this test, each antivirus program had it’s own dedicated machine which was connected to the internet and was interacting with hundreds of URLs each day.  It was automated (rather than carried out by human subjects). All the antivirus programs were being updated on the morning of every day. The methodology of this test is quite interesting, actually, so make sure to read the original description of it in the downloadable PDF, which you can find via the av-comparatives website on this page.

In a nutshell, this is testing ‘in the field’, like a poll designed to measure the degree of protection these antivirus programs would offer in the real world over a long period (5 months) — as opposed to an artificial test where the programs are subjected to ‘canned’ threats. The results are in the table below:

Real World Protection Test: average from AUG - NOV 2015
Free Programs  Paid Programs Control
Avast AVG Avira Bitdefender ESET Kaspersky Microsoft
Protection Rate (1) 98.8% 99.3% 99.9% 99.9% 98.7% 99.9% 94.5%
Wrongly blocked
domains & files (2)
6 0 6 3 2 0 29
AV Comparatives
Award level (3)
Advanced +
3 stars
Advanced +
3 stars
Advanced +
3 stars
Advanced +
3 stars
Advanced +
3 stars
Advanced +
3 stars
Standard
1 star
(1) Higher is better. The average real world protection rate for all was 97.53%. (2) Lower is better. The average wrongly blocked domains and files for all programs tested was 15. (3) AV Comparatives suggest that products with the same award be considered equivalent in their real world protection effectiveness.

 

Interpretation: what we can conclude from the “Real World Protection Test AUG – NOV 2015”:

  1. The three best free Antivirus programs (Avira, AVG, and Avast) and the three best paid ones (Kaspersky, BitDefender and ESET) are all comparable in their protection rate and the rate of wrongly blocked domains and files, in a real world test over five months, despite minuscule yet ultimately insignificant variations.
  2. The six programs above are all significantly better than the control (Microsoft Security Essentials).

 

2. Performance Test (OCT 2015)

A test designed to measure the extent to which the antivirus program uses up system resources and/or contributes to slowing down your system, especially during tasks that antivirus software typically monitors. Each program was running on its own dedicated PC and a benchmarking app PC Mark 8 Professional used to measure system performance. The tasks involved in the test were: file copying, archiving and archiving, installing and uninstalling applications, launching applications, and downloading files.

If you’d like to know more about this test, you can download the detailed PDF from this page. The results are displayed in the table below:

Performance Test: OCT 2015
Free Programs  Paid Programs Control
Avast AVG Avira Bitdefender ESET Kaspersky Microsoft
System Impact Score (4) 2.7 7.2 2.3 7.5 7.7 5.4 16.2
AV Comparatives
Award level (5)
Advanced +
3 stars
Advanced +
3 stars
Advanced +
3 stars
Advanced +
3 stars
Advanced +
3 stars
Advanced +
3 stars
Advanced
2 stars
(4) Lower is better. The average system impact score  in this test was 13.07.  (5) AV Comparatives suggest that products with the same award be considered equivalent in their system resource usage.

 

Interpretation: what we can conclude from the “Performance Test Oct 2015” test:

  1. The three best free Antivirus programs (Avira, AVG, and Avast) and the three best paid ones (Kaspersky, BitDefender and ESET) are all comparable in the degree that they impact the system or use up resources, despite minuscule (yet ultimately insignificant) variations.
  2. The six programs above are all significantly better than the control (Microsoft Security Essentials), in terms of resource usage.
  3. The three free programs are generally lighter than the paid programs, with Avira and Avast taxing the system least of all, although not significantly so.

 

3. File Detection and False Positives Test – (September 2015)

This is a more typical test that the ‘real world’ test above, in that products were essentially subjected to a collection of malware (166522 malware samples), and the detection rate measured. Note that the real world test is probably more relevant to real world users who get exposed to malware from the internet; however, this one is a good test for detection of malware that might come from sources other than the internet or for the best antivirus options to use on an already infected system.

If you’d like to know more about this test, you can download the detailed PDF from this page. The results are displayed in the table below:

Malicious File Detection and False Positives Test: SEP 2015
Free Programs  Paid Programs Control
Avast AVG Avira Bitdefender ESET Kaspersky Microsoft
Detection Rate (6) 99.2% 93.4% 99.8% 99.8% 99.2% 99.5% 91.4%
False Positives (7) 35 6 4 2 0 2 0
AV Comparatives
Award level (8)
Advanced
2 stars
Standard
1 star
Advanced +
3 stars
Advanced +
3 stars
Advanced +
3 stars
Advanced +
3 stars
Tested
No star
(6) Higher is better. The average detection rate for all in this test was xx.xx%. (7) Lower is better. The average for programs tested was 15. (8) AV Comparatives suggest that products with the same award be considered equivalent in their correct detection rates.

 

Interpretation: what we can conclude from the “File Detection and False Positives Test – (September 2015)” test:

  • All the paid programs did really well on this test and had high detection rates and low false positives.
  • As for the free options: Avira is a cut above the rest, comparable to the paid programs. AVG had a comparatively poor detection rate, while the good detection rate for Avast was dragged down by a very high rate of false positive.
  • The six programs above are all significantly better than the control (Microsoft Security Essentials), in terms of detection rate.
  • However, be mindful that this is a test within an artificial environment that does not exist in the real world. Users in the real world will experience the kind of protection illustrated in test #1 above.

 

The Verdict: 

Summary of Tests
Free Programs  Paid Programs Control
Avast AVG Avira Bitdefender ESET Kaspersky Microsoft
1. Real World Protection
Test AUG – NOV 2015
WEIGHT=50%
Advanced +
3 stars
Advanced +
3 stars
Advanced +
3 stars
Advanced +
3 stars
Advanced +
3 stars
Advanced +
3 stars
Standard
1 star
2. Performance Test:
OCT 2015
WEIGHT=35%
Advanced +
3 stars
Advanced +
3 stars
Advanced +
3 stars
Advanced +
3 stars
Advanced +
3 stars
Advanced +
3 stars
Advanced
2 stars
3. Malicious File
Detection & False
Positives Test: SEP 2015
WEIGHT=15%
Advanced
2 stars
Standard
1 star
Advanced +
3 stars
Advanced +
3 stars
Advanced +
3 stars
Advanced +
3 stars
Tested
No star

Please note: don’t average results for all three tests. Test #1 would weigh 50% in my estimation, followed by 35% weight for test #2, and only 15% for test #3.

Interpretation: to summarize the overall picture

  • It is possible to conclude: all of the 6 antivirus programs mentioned here are essentially excellent, as evidenced by test #1. Also, all of them beat the control significantly (Microsoft Security Essentials)
  • If you insist on finding a champion”best free antivirus”, then Avira is it, providing high detection rates across the board with low false positives, and is lightest on the system’s resources to boot (even when compared to ALL the programs mentioned in this post, free or paid).
  • Avast would be the second best free option, adversely impacted only by a (possible) high rate of false positives, and AVG could be ranked a very close third (the reason: a lower relative detection rate when artificially bombarded with canned malware).
  • Having said the above, I will repeat: the differences between the free options might well be negligible. If you like any one of the three, are used to it, or just happen to have it installed on your system — then stick with it.

Download links:


 
 
 
Flattr this!
  • http://cyber-d.blogspot.com dariovolaric

    I have been using Avast for years now. What I loved about it is it’s clean interface and the fact that it has a sandbox (most other free AV’s don’t). However, just today I had to remove Avast from 2 Windows 10 machines. Yesterday Windows released an update and the machines were stuck in a “Cortana and Startmenu are not starting, please sign out” loop. I read on a forum that uninstalling Avast would make your computer work again. I thought “It can’t be that simple!” and so I tried because I didn’t feel like reinstalling Windows. Opened a browser windows, downloaded and installed “geek uninstaller” and uninstalled Avast, then rebooted. And it worked! I could use my computer again! Now I am looking into Comodo again (been a Comodo fan before they became bloatware and moved to Avast).

  • AJNorth

    Since September of 2008, Avira (free) has been my A-V of choice for the various machines I own or maintain, from Win XP through 8.1; now with an aggregate of well over one hundred machine-years, to date none have experienced an infection (in spite of some of their users…).

    Its configurability allows Avira to be dialed-in for each
    installation, though the default protection should be adequate for most
    users.

    In the interest of full disclosure, additional layers of protection are employed, including Firefox (with NoScript, HTTPS-Everywhere, the Web Of Trust and, more recently, uBlock Origin and Privacy Badger), EMET, WinPatrol and SpywareBlaster; the Secunia PSI is also installed. Following best practices, regular on-demand scans are performed with Malwarebytes, AdwCleaner and HitmanPro to improve confidence. The DNS for each is set to OpenDNS.

    In sum, Avira continues to prove itself to be an excellent free antimalware solution.

  • Robert Tipping

    Great insight as usual hank you so much for the time and care you put into your artcles.

  • http://blog.danielmitran.ro/ Daniel Mitran

    Bitdefender free edition; is there anything wrong with it so that it’s out of the free list?

    • http://nerotic.net/ nerotic

      I have the same question. In the same comparatives their scores are above Avast and AVG.

      • SamerKurdi

        @nerotic: see my respone to Daniel above.

    • SamerKurdi

      You know, it’s embarrassing, but I actually didn’t know that Bitdefender had a free edition. Obviously the Bitdefender engine according to the tests included here would be on par with Avira and share the “first prize” so to speak. I am going to test it out on my machine soon just for curiosity, and tweak the article to accordingly.

    • MountainKing7

      You beat me to it Daniel. The Bitdefender free is an absolute gem. Unobstrusive, lightweight and catches so much :)

    • marc

      are you sure? I tried to download this free edition on their website bit I was just clicking around in circles…. is it posssible that they stopped offering a free version? All I find is a 30-days trial version.

    • http://carbonize.co.uk/ Carbonize

      Well unless something has changed since I used it Bitdefender Free comes with a total lack of control. If it finds a file it doesn’t like it just gets rid of it, no warning, no asking what you want to do, just gone. OK for a virus, not so good for a false positive.

  • Stephen Fletcher

    Sophos now do a free home edition…
    https://www.sophos.com/en-us/lp/sophos-home.aspx

    • SamerKurdi

      Thanks for letting me know about this Stephen. I just went back and looked and it looks like Sophos underperfomed the other free AV software in the tests presented above, so it would most likely be a ‘second tier’ option. In any case I may add it to the article in a future update.

  • victor

    Why is MS Security Essentials (Now simply called Defender in Windows 10) consistently ranked lower than other AV / Malware products? It has essentially the same protection as Forefront the MS Enterprise security package. Corporate users pay $ thousands for that one. Are their IT people simply wrong?

    Do the other AV companies target MS as the test case and tweak their packages accordingly? After all as the built-in competition MS is the elephant in the room.

    This review would be more credible if you could reach out to Microsoft and ask them the same questions.

    • SamerKurdi

      I don’t know the answer to these questions. But I do know one thing: MS Security Essentials / Windows defender under-performed significantly in all of the above tests. I am guessing that, yes, many of the other AV programs make it a point to be better than the MS software. That is in fact the logic of using the MS program as the control state.

  • http://carbonize.co.uk/ Carbonize

    You cannot use av-comparatives to judge the free versions since they always use the full versions for their testing so the results may not be the same.

    At present I am using Total Security Essentials from Qihoo. It comes with their own, slightly flawed, cloud engine (occasional false positive) but it also can run Avira and Bitdefender engines side by side with their own.