Lines of code on a dark monitor

Antivirus Test Results 2026: AV-TEST and AV-Comparatives Compared

By Thomas Lovas Lokoy · Published November 7, 2026 — researched and drafted with AI assistance from published sources; edited and published by NorwegianSpark SA.

Which antivirus tested best in 2026? On the results AV-TEST and AV-Comparatives had published by November 2026, Bitdefender and Kaspersky had the strongest all-round records: full marks in every AV-TEST Windows round, the top two protection rates in AV-Comparatives' real-world test, and the top ADVANCED+ award in its September malware and performance tests. Norton and Avast, which share one engine, were a step behind only on false alarms. Microsoft Defender, the free protection built into Windows, scored well on protection but raised more false alarms and slowed the test PC more than the leaders. Every figure below was read from the two labs' own results pages on 7 November 2026.

Lab results answer one question, how well a product stops malware without getting in the way, and they answer it well. They do not answer whether you should trust a vendor, which features you need or what it costs. Those questions are covered in our best antivirus guide; this page is the evidence underneath it.

The 2026 results at a glance

ProductAV-TEST Windows, Jan-Aug 2026AV-Comparatives real-world, Feb-MayAV-Comparatives malware, SepFalse alarms, Sep
Bitdefender Total Security18/18 in all four rounds99.5%99.98%, ADVANCED+1
Kaspersky Premium18/18 in all four rounds99.8%99.97%, ADVANCED+5
Norton (360 / AntiVirus Plus)18/18 in all four rounds99.3%99.96%, ADVANCED15
Avast Free Antivirus18/18 in all four rounds99.3%99.96%, ADVANCED15
McAfee Total Protection18/18 in all four rounds98.5%99.97%, STANDARD36
ESET17.5, 17.5, 18, 1898.5%99.89%, ADVANCED+3
Microsoft Defender18, 18, 17.5, 17.599.0%99.89%, STANDARD28
Malwarebytes Premium17.5 (Mar-Apr), 18 (Jul-Aug)98.8%99.65%, TESTED32

AV-TEST totals are protection plus performance plus usability, 6 points each; the four rounds are January to February, March to April, May to June and July to August. Malwarebytes was not in AV-TEST's other two rounds. AV-Comparatives' real-world figure is the share of 400 live web attacks blocked; its malware figure is the online protection rate against 10,000 malicious files.

How to read the two labs

AV-TEST scores each product out of 6 for protection, performance and usability, and says it awards its certificate at 10 points or more and its TOP PRODUCT award at 17.5 points or more. In practice the top of the field is crowded: in the July to August 2026 round, 14 of the 15 products scored the full 6 for protection. AV-TEST is good at telling you that a product is in the top tier. It is not built to rank the products within it.

AV-Comparatives runs separate tests, publishes raw counts, and gives ranked awards in which, in its own words, "the number of false positives can also affect a product's rating". That is why a product can block almost everything and still drop a level, as McAfee and Microsoft Defender did in September. The lab itself warns against "purchasing a product purely on the basis of one individual test or even one type of test", which is the reason this page puts several side by side.

A false alarm is a clean file or website wrongly blocked. A handful a year is a nuisance; a product that raises them often trains people to click through warnings, which is a risk of its own.

AV-Comparatives Real-World Protection Test, February to May 2026

This is the closest lab test to ordinary web browsing: 400 live malicious sites and downloads, run on Windows with each product connected to its cloud service.

ProductsBlocked of 400Protection rate
Kaspersky39999.8%
Bitdefender39899.5%
Avast, AVG, Norton39799.3%
Microsoft, TotalAV39699.0%
F-Secure, Malwarebytes39598.8%
ESET, K7, McAfee39498.5%

The whole top half of the table sits within five test cases of each other. Treat the order as a snapshot rather than a ranking that will hold next time.

AV-Comparatives Malware Protection Test, September 2026

Here malicious files are run from sources such as USB drives and network shares rather than the web. The test set held 10,000 samples. Offline detection is what a product recognises with no internet connection; the protection rate is the final result with its cloud service available.

ProductOffline detectionProtection rateFalse alarmsAward
Bitdefender97.2%99.98%1ADVANCED+
Kaspersky89.8%99.97%5ADVANCED+
ESET95.7%99.89%3ADVANCED+
Avast95.4%99.96%15ADVANCED
Norton95.4%99.96%15ADVANCED
McAfee81.3%99.97%36STANDARD
Microsoft Defender82.2%99.89%28STANDARD
Malwarebytes84.2%99.65%32TESTED

The March 2026 round of the same test showed a similar shape, with fewer false alarms overall: Kaspersky 2, Microsoft 3, Bitdefender 4, Avast and Norton 9 each, McAfee 14 and Malwarebytes 23. Two products stand out across both rounds. Bitdefender combined top protection with very few false alarms. Kaspersky's offline detection was the lowest of the leaders, so it leans on its online service more than they do.

AV-Comparatives Performance Test, September 2026

How much each product slowed a Windows 11 PC during everyday tasks. Lower impact scores are better.

ProductImpact scoreAward
McAfee3.1ADVANCED+
Kaspersky3.3ADVANCED+
Norton4.9ADVANCED+
Avast5.2ADVANCED+
Bitdefender8.7ADVANCED+
ESET9.5ADVANCED+
Microsoft Defender13.5ADVANCED
Malwarebytes19.1ADVANCED

AV-Comparatives notes that this test measures speed only and says nothing about protection. Anything with ADVANCED+ will feel light on a modern PC; the gaps matter most on older or low-powered machines.

Mac results in 2026

AV-TEST's June 2026 round on macOS Tahoe tested 9 products, and Avast One, Bitdefender Antivirus for Mac, Kaspersky Premium and Norton 360 each scored 6/6/6. AV-Comparatives' Mac test for April to May 2026 used 1,500 Mac malware samples: Kaspersky Premium caught 100%, Avast One Free, AVG and Norton AntiVirus Plus 99.9%, Bitdefender 99.0% and Intego ONE Complete 96.7%, and none of the products raised a false alarm. Whether a Mac needs any of this is a separate question, answered in our best antivirus for Mac guide.

Why some products score the same

Several brands run another company's detection engine, and AV-Comparatives lists them. In its September 2026 malware test, AVG and Norton used the Avast engine; G Data, Total Defense and VIPRE used the Bitdefender engine; and F-Secure, Fortect and TotalAV used the Avira engine. Its Mac report adds that Avast, AVG and Norton are all products of Gen Digital. That is why Avast and Norton keep posting identical numbers, and why choosing between them is about the bundle rather than the protection; see Avast vs Norton.

Brands people search for that the labs did not test

Which result should decide your choice?

For two common head-to-heads on this data, see Bitdefender vs Avast and Bitdefender vs Norton. For on-demand scanners as a second opinion, see best malware scanners, and for how to weigh features and price, the antivirus buying guide.

We earn a commission from some links on this site, and it can influence which providers we feature; it does not change the lab figures above, which are copied from the testers' own pages. AV-TEST published a new Windows round every two months in 2026, so check both labs' pages for anything published after 7 November 2026. This is general guidance, not security advice for a specific threat model.

Frequently Asked Questions

Which antivirus has the best test results in 2026?
Bitdefender and Kaspersky had the strongest all-round records in the 2026 results published by AV-TEST and AV-Comparatives: 18 out of 18 in every AV-TEST Windows round from January to August, the two highest protection rates in AV-Comparatives' February to May real-world test (Kaspersky 99.8%, Bitdefender 99.5%), and ADVANCED+ in its September malware and performance tests. Bitdefender raised the fewest false alarms, 1 in September.
How did Norton and Avast do in the 2026 antivirus tests?
Identically, because AV-Comparatives lists Norton as using the Avast engine. Both scored 18 out of 18 in every AV-TEST Windows round from January to August 2026, blocked 397 of 400 cases (99.3%) in AV-Comparatives' real-world test, and reached a 99.96% protection rate with 15 false alarms in its September malware test.
Is Microsoft Defender good enough according to the 2026 tests?
On protection it was close to the leaders: 99.0% in AV-Comparatives' February to May real-world test and 6 out of 6 for protection in every AV-TEST round. It raised 28 false alarms in AV-Comparatives' September malware test and had an impact score of 13.5 in the September performance test, both weaker than the top paid products.
Is Symantec antivirus tested in 2026?
Not under that name. No Symantec consumer product appears in AV-TEST's or AV-Comparatives' 2026 home-user tests. The company sold its enterprise business and the Symantec brand to Broadcom in November 2019 and renamed itself NortonLifeLock, so its consumer antivirus is tested as Norton.
Why do AV-TEST and AV-Comparatives give different results?
They test different things in different ways. AV-TEST scores protection, performance and usability out of 6 each and most leading products reach the maximum. AV-Comparatives runs separate real-world, malware and performance tests, publishes raw counts and lowers awards for false alarms, so it separates the top products more clearly.

Affiliate disclosure

This article contains affiliate links. If you purchase through them, CyberTechVault earns a commission at no extra cost to you. Our assessments are based on vendors' published documentation, independent lab results and security disclosures — not on hands-on testing by us. We earn a commission from some links on this site; it can influence which providers we feature.

Full disclosure: /affiliate-disclosure.

Sources

Factual claims above were checked against these primary sources. Verify directly on the source for anything time-sensitive before relying on it.

Continue reading

Continue in this collection