The Companies and Products

LoadStorm Testing Results

The Load Storm test (in k6) is designed to simulate real users visiting the site, logging in and browsing. It tests uncached performance.

The key metrics are:

• Total Requests - Number of requests k6 made
• Total Errors - Number of error requests
• p95 - The 95th percentile response time
• Page Cum Avg - Cumulative average response time for Pages (html)
• Asset Cum Avg - Cumulative average response time for Assets (css, js, images)
• Login Cum Avg - Cumulative average response time for Login (wp-profile.php)
• WP-Login Avg Response Time - Average response time for login (301 redirect which validates credentials)

Results Table

Discussion

Pressable, Presslabs, WordPress.com, and WPX all finished the LoadStorm test with zero errors and flat response times. The per-persona breakdown (below) is where the differences between them show up.

GreenGeeks handled the test with only 5 errors against 6.6M requests. Cart and login averages crossed 1 second at multiple points, so the test wasn't as flat as the four no-issue results above, but the server kept delivering through the run.

HostArmada was ok until around 600 VUsers when an error spike kicked off and response times started climbing for the rest of the run.

Seravo started having trouble around 700 VUsers and finished the test with 1.55M errors against 12M requests.

Trustdom looked great up to about 800 VUsers, then a security or rate-limiting layer cut in and started blocking everything — by the end of the test 6.36M of 8.52M requests had errored.

Load Storm Average Response Time by Profile

This takes deeper look at the results and analyzes the performance of loading HTML mime types, to ignore a lot of static assets which are cached and bring down average response times. This tests how fast the initial pages are delivered, which is what a user would experience before loading all the other assets like css, javascript and images.

Results Table

Pressable was in the top three on four of the five personas (Cart 240ms third, Browser 7ms second, Buyer 104ms second, Homepage 8ms second) and fourth on Customer (138ms) by a 7ms margin. Consistent across the board.

Presslabs had the fastest Buyer (72ms) and Customer (89ms) profiles, and tied Trustdom for the fastest Cart at 207ms. Three of the five fastest persona times.

WPX had the fastest Browser (5ms) and Homepage (5ms) — their static-cache layer is fast — but the Cart slowed to 523ms. The gap kept WPX at Honorable Mention.

WordPress.com was top three on Browser (10ms), Homepage (14ms), and Customer (131ms). Cart at 407ms was middle of the pack.

GreenGeeks was middle-of-pack on every persona — never the fastest, never the slowest.

HostArmada was the slowest on Cart (2721ms), Buyer (1433ms), and Customer (1457ms). Seravo was the slowest on Browser (289ms) and Homepage (273ms), and second-slowest on the other three (Cart 1335ms, Buyer 783ms, Customer 872ms).

Trustdom's persona numbers look strong (Cart tied for fastest at 207ms, Buyer 126ms, Customer 125ms), but the LoadStorm test got cut off by the security layer mentioned above before reaching peak load. The per-persona Cumavg values cover the portion of the test that completed cleanly, not the full ramp.

K6 Static Testing Results

K6 Static test is designed to test cached performance by repeatedly requesting the homepage.

Results Table

Discussion

Pressable had the fastest average response time at 5.6ms, WordPress.com was second at 7.6ms, and WPX was third at 8.0ms — all three under 10ms on the static test.

Presslabs (88.9ms), HostArmada (112.4ms), and GreenGeeks (119.4ms) sit in the next cluster, with Seravo (210.6ms) behind them.

Trustdom's 992.5ms average with 2,761 errors lines up with the same security cut-in seen on the dynamic test — once the rate-limiting layer kicked in, the static test slowed and errors started arriving.

Uptime Testing Results

Uptime is monitored by three sources: HetrixTools and StatusCake as third-party services, plus a self-hosted UptimeKuma instance for cross-reference in case of major discrepancy between the third-party monitors.

Results Table

Discussion

Pressable had 100% across all three monitors. WordPress.com was near perfect (99.9987% HT, 100% SC, 99.9959% UK). Trustdom cleared 99.9% on all three.

Presslabs is the same single-monitor outlier as the enterprise tier — HetrixTools 100% and UptimeKuma 99.9941%, but StatusCake came in at 99.38%. Top Tier stands; the StatusCake number is what's in the table.

WPX came in below 99.9% on StatusCake (99.83%) and UptimeKuma (99.78%). The 99.78% UptimeKuma value is the same one their WordPress results showed at $25, $50, and $100, which kept this result at Honorable Mention rather than Top Tier.

Seravo came in at 99.87% on StatusCake while the other two monitors had a clean year. GreenGeeks and HostArmada both came in below 99.9% on multiple monitors — GreenGeeks on all three (99.61% / 99.39% / 99.72%) and HostArmada on two (99.66% HT, 99.83% UK). Those are real uptime issues and contributed to keeping both out of awards.

GTmetrix Testing Results

GTmetrix loads the homepage and records how long it takes from 12 different locations around the world. (We switched from WebPageTest to GTmetrix in 2026.) Results are measured in seconds.

Results Table

Discussion

Pressable had the fastest average and was the fastest in 5 of the 12 locations (San Antonio TX, Cape Town, Mumbai, Brazil, and Dubai). WordPress.com had the second fastest average and won 4 locations (California, Singapore, Tokyo, and Sydney). GreenGeeks had the third fastest average and was the fastest in Virginia and Frankfurt. Trustdom was the fastest in London — their European peering looked good there.

WPX's Mumbai measurement above 5 seconds is more likely a routing event than a steady-state result and skews their average. The chart can be adjusted by clicking on a company to filter outliers.

WPPerformanceTester Testing Results

WPPerformanceTester performs two benchmarks. One is a WordPress (WP Bench) and the other is a PHP Bench. WP Bench measures how many WP queries per second and higher tends to be better (varies considerably by architecture). PHP Bench performs a lot of computational and some database operations which are measured in seconds to complete. Lower PHP Bench is better.

Results Table

Discussion

Trustdom had the fastest PHP Bench at 7.91s. Pressable was second at 8.10s and WPX was third at 8.15s — the top three were very close.

The WP Bench results were spread more widely. GreenGeeks had the fastest WP Bench. Presslabs was second and WordPress.com was third.

WPBenchmark Score

WPBenchmark is a hosting performance benchmark that produces a single composite score per server. Higher is better.

Results Table

Discussion

Pressable led the WPBenchmark.io scores at 8.3. Presslabs, WordPress.com, and WPX all tied at 7.1 just behind. HostArmada‘s 2.6 was the lowest score recorded at any tier this year.

SSL Testing Results

The tool is available at https://www.ssllabs.com/ssltest/

Results Table

Discussion

No B or A- grades at this tier this year.

HostArmada, Presslabs, and WPX earned A.

GreenGeeks, Pressable, Seravo, Trustdom, and WordPress.com earned A+.

Conclusion

There are two levels of recognition awarded to companies that participate in the tests. There is no ‘best’ declared, it’s simply tiered, it’s hard to come up with an objective ranking system because of the complex nature of hosting. These tests also don’t take into account outside factors such as reviews, support, and features. It is simply testing performance as described in the methodology.

Top Tier

This year's Top Tier WordPress Hosting Performance Award goes to the following companies who showed virtually no signs of struggle during the testing.

Honorable Mention

The following companies earned Honorable Mention status because they did very well and had a minor issue or two holding them back from earning Top Tier status.

Individual Host Analysis

The Companies and Products

LoadStorm Testing Results

The Load Storm test (in k6) is designed to simulate real users visiting the site, logging in and browsing. It tests uncached performance.

The key metrics are:

• Total Requests - Number of requests k6 made
• Total Errors - Number of error requests
• p95 - The 95th percentile response time
• Page Cum Avg - Cumulative average response time for Pages (html)
• Asset Cum Avg - Cumulative average response time for Assets (css, js, images)
• Login Cum Avg - Cumulative average response time for Login (wp-profile.php)
• WP-Login Avg Response Time - Average response time for login (301 redirect which validates credentials)

Results Table

Discussion

Pressable, WordPress VIP, and Presslabs all had flat response times across the duration of the test. Presslabs and WordPress VIP both have a visible spike at the very end which is a testing-cleanup artifact — by that point RPS has dropped to a trickle of the slowest still-running queries and that inflates the average without representing real degradation.

Altis Cloud's graph is the most interesting at this tier. Response times rose slightly as the load ramped, then dropped as the test continued, their managed autoscaling kicked in and improved performance through the second half of the run. 219 errors against 3.8M requests is tiny.

BigScoots had a strong enterprise LoadStorm test. The aggregate response time actually trended downwards as the test scaled, pageload times got faster under load. Asset and login response times stayed flat, pageloads improved. 345 errors against 4.27M requests is miniscule.

Seravo hit a problem we weren't able to diagnose or resolve before the testing window closed. Errors started piling up early and the test finished with 2.2M errors against 5.79M requests, so the result couldn't be awarded.

K6 Static Testing Results

K6 Static test is designed to test cached performance by repeatedly requesting the homepage.

Results Table

Discussion

Pressable had the fastest average response time at 5.6ms, WordPress VIP was second at 6.0ms, and Altis Cloud was third at 8.9ms. Presslabs (21.7ms) and BigScoots (38.2ms) round out the awarded field, both still excellent. The spike on the Presslabs graph is to 44ms. The spike on the BigScoots graph is a single Singapore blip that moved the average by a couple ms.

Seravo's 134ms average with no errors went fine on the static side — the issues were specific to LoadStorm, not the underlying static performance.

Uptime Testing Results

Uptime is monitored by three sources: HetrixTools and StatusCake as third-party services, plus a self-hosted UptimeKuma instance for cross-reference in case of major discrepancy between the third-party monitors.

Results Table

Discussion

WordPress VIP had 100% across all three monitors. BigScoots hit 100% on HetrixTools and StatusCake with 99.9924% on UptimeKuma. Pressable hit 100% on the two third-party monitors and 99.9994% on UptimeKuma. Altis Cloud hit 100% on StatusCake and was above 99.99% on the other two. Seravo cleared 99.9% on all three.

Presslabs is the one to flag. HetrixTools recorded 100% and UptimeKuma 99.9954%, but StatusCake came in at 99.01%. Looks like a monitor issue when the other two say everything was fine. The Top Tier award stands but the StatusCake number is what's in the table.

GTmetrix Testing Results

GTmetrix loads the homepage and records how long it takes from 12 different locations around the world. (We switched from WebPageTest to GTmetrix in 2026.) Results are measured in seconds.

Results Table

Discussion

BigScoots had the fastest WPT average by a significant margin and was the fastest in 9 of the 12 locations (California, London, Frankfurt, Cape Town, Singapore, Mumbai, Tokyo, Sydney, and Brazil).

Altis Cloud was the fastest in Virginia, Pressable in San Antonio TX, and WordPress VIP in Dubai. Pressable had the second fastest average and WordPress VIP the third.

Seravo had the slowest average at this tier. Their European peering is strong (Frankfurt and London under 500ms) but the Asia-Pacific and Americas legs add up.

WPPerformanceTester Testing Results

WPPerformanceTester performs two benchmarks. One is a WordPress (WP Bench) and the other is a PHP Bench. WP Bench measures how many WP queries per second and higher tends to be better (varies considerably by architecture). PHP Bench performs a lot of computational and some database operations which are measured in seconds to complete. Lower PHP Bench is better.

Results Table

Discussion

BigScoots had the fastest PHP Bench at 5.16s by a significant margin over the rest of the field. Pressable was second at 8.96s and Seravo was third at 9.62s.

BigScoots also had the fastest WP Bench by a significant margin. Presslabs was second and Pressable was third.

WPBenchmark Score

WPBenchmark is a hosting performance benchmark that produces a single composite score per server. Higher is better.

Results Table

Discussion

The WPBenchmark.io scores showed BigScoots led at 9.5, Pressable was second at 8.3, and Presslabs was third at 7.9.

SSL Testing Results

The tool is available at https://www.ssllabs.com/ssltest/

Results Table

Discussion

No B or A- grades at this tier this year.

Altis Cloud, BigScoots, Pressable, Seravo, and WordPress VIP all earned A+.

Presslabs earned A.

Conclusion

There are two levels of recognition awarded to companies that participate in the tests. There is no ‘best’ declared, it’s simply tiered, it’s hard to come up with an objective ranking system because of the complex nature of hosting. These tests also don’t take into account outside factors such as reviews, support, and features. It is simply testing performance as described in the methodology.

Top Tier

This year's Top Tier WordPress Hosting Performance Award goes to the following companies who showed virtually no signs of struggle during the testing.

Honorable Mention

The following companies earned Honorable Mention status because they did very well and had a minor issue or two holding them back from earning Top Tier status.

No company achieved this status.

Individual Host Analysis

The Companies and Products

LoadStorm Testing Results

The Load Storm test (in k6) is designed to simulate real users visiting the site, logging in and browsing. It tests uncached performance.

The key metrics are:

• Total Requests - Number of requests k6 made
• Total Errors - Number of error requests
• p95 - The 95th percentile response time
• Page Cum Avg - Cumulative average response time for Pages (html)
• Asset Cum Avg - Cumulative average response time for Assets (css, js, images)
• Login Cum Avg - Cumulative average response time for Login (wp-profile.php)
• WP-Login Avg Response Time - Average response time for login (301 redirect which validates credentials)

Results Table

Discussion

All four companies handled the LoadStorm test cleanly. Bluehost (0 errors) and Raidboxes (0 errors) were flat across the run. Pressable had one tiny spike in an otherwise excellent test (9 errors against ~1.08M requests). Tangible kept response times flat through the test (2 errors).

This is the cleanest WordPress LoadStorm tier of the year — no security cut-offs, no rate-limit walls, no capacity ceilings hit on the dynamic test for any of the four.

K6 Static Testing Results

K6 Static test is designed to test cached performance by repeatedly requesting the homepage.

Results Table

Discussion

Pressable had the fastest average response time at 5.1ms, Bluehost was second at 5.7ms, and Tangible was third at 10.8ms. All three turned in excellent static tests with flat graphs.

Raidboxes is the odd one here. Response times stayed constant for served requests, but a rate limit was hit during the static run and over 70,000 HTTP 429s arrived — the same throttling pattern that would have been the story on the dynamic test if it had run longer at higher VUser counts. The 119ms average looks fine on its own, but the error count kept this result out of an award.

Uptime Testing Results

Uptime is monitored by three sources: HetrixTools and StatusCake as third-party services, plus a self-hosted UptimeKuma instance for cross-reference in case of major discrepancy between the third-party monitors.

Results Table

Discussion

Pressable had 100% across all three monitors. Raidboxes was near perfect (100% HT, 100% SC, 99.9994% UK). Bluehost hit 100% on HetrixTools and was above 99.96% on the other two. Tangible cleared 99.9% on all three.

Nothing to flag at this tier. The whole field cleared 99.9% uptime.

GTmetrix Testing Results

GTmetrix loads the homepage and records how long it takes from 12 different locations around the world. (We switched from WebPageTest to GTmetrix in 2026.) Results are measured in seconds.

Results Table

Discussion

Bluehost had the fastest average and was the fastest in 6 of the 12 locations (California, London, Singapore, Tokyo, Brazil, and Dubai). Pressable had the second fastest average and won 4 locations (Virginia, San Antonio TX, Mumbai, and Sydney). Raidboxes was the fastest in Frankfurt and Cape Town — the European peering shows again.

Tangible had the slowest average at this tier; their geographic spread is heavier on the higher-latency regions.

WPPerformanceTester Testing Results

WPPerformanceTester performs two benchmarks. One is a WordPress (WP Bench) and the other is a PHP Bench. WP Bench measures how many WP queries per second and higher tends to be better (varies considerably by architecture). PHP Bench performs a lot of computational and some database operations which are measured in seconds to complete. Lower PHP Bench is better.

Results Table

Discussion

Tangible had the fastest PHP Bench at 6.97s. Raidboxes was second at 7.16s and Pressable was third at 8.10s.

Tangible also had the fastest WP Bench by a significant margin. Raidboxes was second and Bluehost was third.

WPBenchmark Score

WPBenchmark is a hosting performance benchmark that produces a single composite score per server. Higher is better.

Results Table

Discussion

The WPBenchmark.io scores agreed at the top — Tangible led at 9.8, with Bluehost and Pressable tied for second at 8.3.

SSL Testing Results

The tool is available at https://www.ssllabs.com/ssltest/

Results Table

Discussion

Everyone earned A+. Nothing else to report.

Conclusion

There are two levels of recognition awarded to companies that participate in the tests. There is no ‘best’ declared, it’s simply tiered, it’s hard to come up with an objective ranking system because of the complex nature of hosting. These tests also don’t take into account outside factors such as reviews, support, and features. It is simply testing performance as described in the methodology.

Top Tier

This year's Top Tier WordPress Hosting Performance Award goes to the following companies who showed virtually no signs of struggle during the testing.

Honorable Mention

The following companies earned Honorable Mention status because they did very well and had a minor issue or two holding them back from earning Top Tier status.

No company achieved this status.

Individual Host Analysis

The Companies and Products

LoadStorm Testing Results

The Load Storm test (in k6) is designed to simulate real users visiting the site, logging in and browsing. It tests uncached performance.

The key metrics are:

• Total Requests - Number of requests k6 made
• Total Errors - Number of error requests
• p95 - The 95th percentile response time
• Page Cum Avg - Cumulative average response time for Pages (html)
• Asset Cum Avg - Cumulative average response time for Assets (css, js, images)
• Login Cum Avg - Cumulative average response time for Login (wp-profile.php)
• WP-Login Avg Response Time - Average response time for login (301 redirect which validates credentials)

Results Table

Discussion

Bluehost, Pressable, Tangible, WP Buzz, and WPX all handled the LoadStorm test without issue. Bluehost had flat response times and zero errors. Pressable had one spike in an otherwise excellent run. WP Buzz showed the now-familiar cron-related bumps which didn't affect capacity.

InMotion Hosting kept response times in check until around 700 VUsers, after which response times started climbing for the rest of the test. The error count stayed low (283) but the slow-down was clear.

Raidboxes looked good up to about 1,000 VUsers, then started slowing down with HTTP 429s ramping up around 1,250 VUsers indicating a rate limit. The error total ran into the hundreds of thousands by the end, which kept the otherwise strong performance up to that ceiling out of an award.

K6 Static Testing Results

K6 Static test is designed to test cached performance by repeatedly requesting the homepage.

Results Table

Discussion

WPX had the fastest average response time at 4.1ms. Pressable was second at 7.2ms and Bluehost was third at 8.4ms. WP Buzz (8.6ms) and Tangible (12.9ms) round out the sub-15ms group.

Bluehost's static graph has a step-pattern partway through that looks substantial in isolation but is a sub-3ms shift in absolute terms — looks like a routing change for the Bangalore source mid-test. Easy to forgive when the axis labels are that small.

InMotion Hosting's 119.5ms average puts them in the no-edge-cache cluster. The graph shows the same 5-minute cron pattern visible on their dynamic test.

Raidboxes had the same rate-limit story on the static test — response times held steady but the rate limit cut in around 650 VUsers and 429 errors started accumulating around 1,750 VUsers, pulling the average to 854ms by the end.

Uptime Testing Results

Uptime is monitored by three sources: HetrixTools and StatusCake as third-party services, plus a self-hosted UptimeKuma instance for cross-reference in case of major discrepancy between the third-party monitors.

Results Table

Discussion

Pressable had 100% across all three monitors. Bluehost, WP Buzz, and InMotion Hosting all hit 100% on at least one monitor and were above 99.99% on the others.

WPX came in at 99.78% on the self-hosted UptimeKuma instance — the same value their lower-tier results show, which points to an UptimeKuma-side issue with that endpoint rather than three independent outages. HetrixTools (100%) and StatusCake (99.9%) on the same plan both had a normal year.

Everyone else was 99.9%+ across all three monitors.

GTmetrix Testing Results

GTmetrix loads the homepage and records how long it takes from 12 different locations around the world. (We switched from WebPageTest to GTmetrix in 2026.) Results are measured in seconds.

Results Table

Discussion

WP Buzz had the fastest average and was the fastest in 11 of the 12 locations. Raidboxes was the fastest in Frankfurt — their European peering shows. Pressable had the second fastest average and Bluehost the third, both without winning an individual location.

The slowest averages were WPX, Tangible, and InMotion Hosting. The WPX number is pulled up by a single Mumbai measurement above 8 seconds that looks like a routing event, not a steady-state result.

WPPerformanceTester Testing Results

WPPerformanceTester performs two benchmarks. One is a WordPress (WP Bench) and the other is a PHP Bench. WP Bench measures how many WP queries per second and higher tends to be better (varies considerably by architecture). PHP Bench performs a lot of computational and some database operations which are measured in seconds to complete. Lower PHP Bench is better.

Results Table

Discussion

WP Buzz had the fastest PHP Bench at 4.86s. Tangible was a fraction behind at 4.93s. WPX was third at 7.22s.

Tangible had the fastest WP Bench by a significant margin. WP Buzz was second and Bluehost was third.

WPBenchmark Score

WPBenchmark is a hosting performance benchmark that produces a single composite score per server. Higher is better.

Results Table

Discussion

The WPBenchmark.io scores were very close at the top — WP Buzz and Tangible tied at 9.9, with Bluehost and Pressable tied for third at 8.3.

SSL Testing Results

The tool is available at https://www.ssllabs.com/ssltest/

Results Table

Discussion

No B or A- grades at this tier this year.

Bluehost, Pressable, Raidboxes, and Tangible earned A+.

InMotion Hosting, WP Buzz, and WPX earned A.

Conclusion

There are two levels of recognition awarded to companies that participate in the tests. There is no ‘best’ declared, it’s simply tiered, it’s hard to come up with an objective ranking system because of the complex nature of hosting. These tests also don’t take into account outside factors such as reviews, support, and features. It is simply testing performance as described in the methodology.

Top Tier

This year's Top Tier WordPress Hosting Performance Award goes to the following companies who showed virtually no signs of struggle during the testing.

Honorable Mention

The following companies earned Honorable Mention status because they did very well and had a minor issue or two holding them back from earning Top Tier status.

No company achieved this status.

Individual Host Analysis

The Companies and Products

LoadStorm Testing Results

The Load Storm test (in k6) is designed to simulate real users visiting the site, logging in and browsing. It tests uncached performance.

The key metrics are:

• Total Requests - Number of requests k6 made
• Total Errors - Number of error requests
• p95 - The 95th percentile response time
• Page Cum Avg - Cumulative average response time for Pages (html)
• Asset Cum Avg - Cumulative average response time for Assets (css, js, images)
• Login Cum Avg - Cumulative average response time for Login (wp-profile.php)
• WP-Login Avg Response Time - Average response time for login (301 redirect which validates credentials)

Results Table

Discussion

Bluehost, GreenGeeks, Pressable, Tangible, WordPress.com, WP Buzz, WPX, and Krystal Hosting all handled the LoadStorm test without issue. Krystal's graph is the most visible cron pattern in the group — two jobs appear to be firing on 1-minute and 5-minute intervals, and when both coincide there's a big spike for a moment, but overall the test was fine.

Arsys Internet also had a clean LoadStorm — zero errors and flat response times with one regional spike. The static test (K6) was a different story (below).

CloudPress had a striking 10-minute spike pattern that was diagnosed to a BunnyCDN security measure that interferes with synthetic testing. We tried but the policy couldn't be exempted for the test. Ignoring those recurring spikes the test looks a lot better, but with the security measure still in play we can't fully separate hardware issues from monitoring interference.

KnownHost showed 5-minute cron-related spikes through the test, and around 850 VUsers the third cron event turned into sustained slow response times for the rest of the run.

InMotion Hosting had a big spike of response times and errors in the middle of the test (~11.6k errors) that the server recovered from, but it cost them the run.

ScalaHosting started slowing down around 350 VUsers with errors arriving around 600 VUsers.

K6 Static Testing Results

K6 Static test is designed to test cached performance by repeatedly requesting the homepage.

Results Table

Discussion

WPX had the fastest average response time at 4.9ms (3.73ms once cached). Pressable was second at 5.0ms and WordPress.com was third at 6.1ms. WP Buzz (7.7ms), Tangible (11.2ms), CloudPress (12.8ms), and Bluehost (15.1ms) were the rest of the sub-30ms group.

The line between companies running edge-cached configurations and those serving from origin is again pretty clear — the next cluster sits in the 100-130ms range and largely matches the no-CDN plans.

Arsys Internet is the interesting one in this section. The average came in at 319ms with 168 errors, and the graph looks like a security/throttle pattern — fine, then not fine, repeating through the test. We've seen this shape before when a per-IP rate limit or resource cap kicks in, drops back, and re-fires.

ScalaHosting had no static cache configured by default and the test slowed down accordingly. The 8.9s average isn't a hardware story, it's an origin-load story.

Uptime Testing Results

Uptime is monitored by three sources: HetrixTools and StatusCake as third-party services, plus a self-hosted UptimeKuma instance for cross-reference in case of major discrepancy between the third-party monitors.

Results Table

Discussion

Pressable was the only result with 100% across all three monitors. WordPress.com, WP Buzz, and InMotion Hosting all stayed above 99.9% on every monitor and were near perfect.

Bluehost is the standout this tier. StatusCake recorded 99.93%, but HetrixTools came in at 94.48% and UptimeKuma at 91.98%. Two monitors registering substantial outages while the third had a normal year is unusual, and the LoadStorm and K6 tests both ran cleanly on the same plan. The WAF probably treated the monitor IPs differently rather than this being a real outage of that size. The honorable mention stands but the numbers as recorded are worth noting.

WPX (SC 99.82%, UK 99.79%), CloudPress (SC 99.86%, UK 99.89%), Krystal Hosting (UK 99.38%), and Arsys Internet (HT 99.88%) each came in below 99.9% on at least one monitor. Everyone else was 99.9%+ across the board.

GTmetrix Testing Results

GTmetrix loads the homepage and records how long it takes from 12 different locations around the world. (We switched from WebPageTest to GTmetrix in 2026.) Results are measured in seconds.

Results Table

Discussion

WP Buzz had the fastest average by a significant margin, picking up 11 of the 12 locations. WordPress.com was the fastest in San Antonio TX. No other company won an individual location.

Pressable had the second fastest average overall and GreenGeeks the third.

Tangible and InMotion Hosting had the slowest averages at this tier. The chart can be adjusted by clicking on a company to filter outliers.

WPPerformanceTester Testing Results

WPPerformanceTester performs two benchmarks. One is a WordPress (WP Bench) and the other is a PHP Bench. WP Bench measures how many WP queries per second and higher tends to be better (varies considerably by architecture). PHP Bench performs a lot of computational and some database operations which are measured in seconds to complete. Lower PHP Bench is better.

Results Table

Discussion

WP Buzz had the fastest PHP Bench at 4.87s. Tangible was right behind at 4.96s and ScalaHosting was third at 6.91s.

Tangible had the fastest WP Bench by a significant margin. Bluehost was second and WP Buzz was third.

WPBenchmark Score

WPBenchmark is a hosting performance benchmark that produces a single composite score per server. Higher is better.

Results Table

Discussion

The WPBenchmark.io scores agreed with the WPPerformanceTester results. Tangible led at 9.9, WP Buzz was second at 9.8, and Bluehost was third at 9.3.

SSL Testing Results

The tool is available at https://www.ssllabs.com/ssltest/

Results Table

Discussion

No B grades at this tier this year.

Arsys Internet earned an A-.

GreenGeeks, Pressable, ScalaHosting, Tangible, and WordPress.com earned A+.

Everyone else earned an A.

Conclusion

There are two levels of recognition awarded to companies that participate in the tests. There is no ‘best’ declared, it’s simply tiered, it’s hard to come up with an objective ranking system because of the complex nature of hosting. These tests also don’t take into account outside factors such as reviews, support, and features. It is simply testing performance as described in the methodology.

Top Tier

This year's Top Tier WordPress Hosting Performance Award goes to the following companies who showed virtually no signs of struggle during the testing.

Honorable Mention

The following companies earned Honorable Mention status because they did very well and had a minor issue or two holding them back from earning Top Tier status.

Individual Host Analysis

The Companies and Products

LoadStorm Testing Results

The Load Storm test (in k6) is designed to simulate real users visiting the site, logging in and browsing. It tests uncached performance.

The key metrics are:

• Total Requests - Number of requests k6 made
• Total Errors - Number of error requests
• p95 - The 95th percentile response time
• Page Cum Avg - Cumulative average response time for Pages (html)
• Asset Cum Avg - Cumulative average response time for Assets (css, js, images)
• Login Cum Avg - Cumulative average response time for Login (wp-profile.php)
• WP-Login Avg Response Time - Average response time for login (301 redirect which validates credentials)

Results Table

Discussion

34SP.com, Bluehost, DreamHost, GoDaddy, GreenGeeks, Heart Internet, ICDSoft, Lightning Base, Pressable, WP Buzz, WPX, and 20i all handled the LoadStorm test without issue. Response times stayed flat for the duration of the run. The spikes in the detailed graphs are small cron-related bumps, not capacity problems.



HostArmada had multiple large spikes through the test where response times jumped substantially before settling back. 9 errors out of 786k requests is fine, so the result still earned recognition with that caveat.



KnownHost handled the test cleanly with no errors but response times started to climb around 650 VUsers.



AleForge started slowing down around 300 VUsers and errors jumped around 500 VUsers. The errors were 429s indicating a rate limit.



FullHost started to struggle around 200 VUsers with errors kicking in around 550 VUsers.



Raidboxes had its first noticeable spike around 600 VUsers and became less stable around 900 VUsers.



Fasthosts and Strato both ran fine for roughly the first 20 minutes and then a security or resource-throttling measure kicked in and the server stopped responding for the remainder of the test. Same shape on both (both IONOS-owned), likely the same platform policy.



mijn.host and Trustdom both had wp-login blocked partway through the test by a security measure that couldn't be worked around.



Zone Media looked flat until a rate-limit/security measure kicked in and from that point on the test was all errors. The actual numbers on the test and the graph are different because of a scripting bug which caused far too many failures than should have happened. The test had 6731820 requests with 6615801 errors, but with the help of AI we estimated the expected number based on what the script should have been doing. Peak and average RPS were also estimated as well to adjust in the table. This doesn't change the result or analysis but does make the results look more reasonable given the size of the test.



one.com and Z.com both struggled The server slowed down and stopped responding to most requests well before the test finished.

K6 Static Testing Results

K6 Static test is designed to test cached performance by repeatedly requesting the homepage.

Results Table

Discussion

Pressable had the fastest average response time at 5.1ms. WPX was second at 5.9ms (the cached portion was 3.73-4.16ms; the average is pulled up by the cold uncached hit at the start). AleForge was third at 6.3ms. WP Buzz (6.9ms), mijn.host (12.2ms), 20i (13.3ms), Bluehost (18.4ms), and GoDaddy (26.3ms) were also sub-30ms.



The line between companies running aggressive edge caching and those serving from origin is pretty clear at this tier — the next cluster sits in the 100-140ms range and largely matches the no-CDN configurations.



Zone Media's static result is a rate-limit story, not a performance story — the test started flat and then the same security measure that broke the dynamic test broke the static one. Trustdom (2.9s) and Z.com (5.8s) had the static test slow before completing. one.com's static came back quickly (217ms average) for the requests that returned, but 75,716 of 77,854 requests errored.

Uptime Testing Results

Uptime is monitored by three sources: HetrixTools and StatusCake as third-party services, plus a self-hosted UptimeKuma instance for cross-reference in case of major discrepancy between the third-party monitors.

Results Table

Discussion

Only three results came in below 99.9% on a single monitor: DreamHost and WPX both posted 99.78% on the self-hosted UptimeKuma instance, and Fasthosts posted 99.88% on HetrixTools. DreamHost and WPX share the exact same UptimeKuma value — the same number across two unrelated hosts almost always points at the monitor rather than the hosts, and both of their other two monitors were above 99.96%. The Fasthosts dip was only on HetrixTools while StatusCake (99.99%) and UptimeKuma (99.98%) had no issues, so it gets the same benefit of the doubt.



Pressable was the only company with 100% across all three monitors. Lightning Base, WP Buzz, WPX, AleForge, 20i, and Heart Internet each hit 100% on at least one monitor. Everyone else was 99.9%+ across all three.

GTmetrix Testing Results

GTmetrix loads the homepage and records how long it takes from 12 different locations around the world. (We switched from WebPageTest to GTmetrix in 2026.) Results are measured in seconds.

Results Table

Discussion

WP Buzz had the fastest average and was the fastest in 8 of the 12 locations (Virginia, London, Singapore, Mumbai, Tokyo, Sydney, Brazil, and Dubai). AleForge had the second fastest average without winning any individual location. Pressable had the third fastest average. GoDaddy was the fastest in the other 4 locations (California, San Antonio TX, Frankfurt, and Cape Town) but came in fourth on the average — California (139ms) and Frankfurt (193ms) are very fast but Mumbai (2000ms) and Brazil (1600ms) pull the mean down.



Trustdom, Z.com, and Heart Internet had the three slowest averages. The chart can be adjusted by clicking on a company to remove outliers.

WPPerformanceTester Testing Results

WPPerformanceTester performs two benchmarks. One is a WordPress (WP Bench) and the other is a PHP Bench. WP Bench measures how many WP queries per second and higher tends to be better (varies considerably by architecture). PHP Bench performs a lot of computational and some database operations which are measured in seconds to complete. Lower PHP Bench is better.

Results Table

Discussion

Heart Internet had the fastest PHP Bench. WP Buzz was second and FullHost was third.

Bluehost had the fastest WP Bench. Heart Internet was second and WP Buzz was third.

WPBenchmark Score

WPBenchmark is a hosting performance benchmark that produces a single composite score per server. Higher is better.

Results Table

Discussion

WP Buzz led the WPBenchmark.io scores with a 9.9. Bluehost was second at 9.5 and Zone Media was third at 9.1.

SSL Testing Results

The tool is available at https://www.ssllabs.com/ssltest/

Results Table

Discussion

No B grades at this tier this year.



34SP.com, Fasthosts, and Strato earned A-.



AleForge, DreamHost, GreenGeeks, HostArmada, Pressable, Raidboxes, and Trustdom earned A+.



Everyone else earned an A.

Conclusion

There are two levels of recognition awarded to companies that participate in the tests. There is no ‘best’ declared, it’s simply tiered, it’s hard to come up with an objective ranking system because of the complex nature of hosting. These tests also don’t take into account outside factors such as reviews, support, and features. It is simply testing performance as described in the methodology.

Top Tier

This year's Top Tier WordPress Hosting Performance Award goes to the following companies who showed virtually no signs of struggle during the testing.

Honorable Mention

The following companies earned Honorable Mention status because they did very well and had a minor issue or two holding them back from earning Top Tier status.

Individual Host Analysis

The Companies and Products

LoadStorm Testing Results

The Load Storm test (in k6) is designed to simulate real users visiting the site, logging in and browsing. It tests uncached performance.

The key metrics are:

• Total Requests - Number of requests k6 made
• Total Errors - Number of error requests
• p95 - The 95th percentile response time
• Page Cum Avg - Cumulative average response time for Pages (html)
• Asset Cum Avg - Cumulative average response time for Assets (css, js, images)
• Login Cum Avg - Cumulative average response time for Login (wp-profile.php)
• WP-Login Avg Response Time - Average response time for login (301 redirect which validates credentials)

Results Table

Discussion

Nexcess, Pressable, and both Servebolt plans handled this test without any issues.

Cloudways started slowing down around 2300 VUsers.

K6 Static Testing Results

K6 Static test is designed to test cached performance by repeatedly requesting the homepage.

Results Table

Discussion

Every company handled the static test almost perfect. There was a decrease in average and p95 response times from every company that participated in the previous year. Cloudways went from 212/489 to 111/217 [Avg/p95]. Nexcess went from 248/473 to 14/22. Pressable went from 5.9/19 to 5.3/12. Servebolt went from 176/283 + 37/50 to 20/28 + 30/41. It shows the companies which were slower last year saw the biggest gains, but even the fastest companies continue to speed up.

Uptime Testing Results

Uptime is monitored by two companies: HetrixTools and Uptime Robot. A self hosted monitor was also run in case there was a major discrepancy between the two third party monitors.

Results Table

Discussion

Perfect uptime on Uptime Robot for everyone and nobody was below 99.99% on Hetrix Tools.

WebPageTest Testing Results

WebPageTest fully loads the homepage and records how long it takes from 12 different locations around the world. Result are measured in seconds.

Results Table

Discussion

Servebolt was the fastest average with 406ms with 3/12 fastest locations.

Servebolt ACD followed close behind for second at 415ms average and 2/12 fastest locations.

Pressable was third fastest average with 451ms with 6/12 fastest locations

Cloudways was the fastest in Virginia.

WPPerformanceTester Testing Results

WPPerformanceTester performs two benchmarks. One is a WordPress (WP Bench) and the other is a PHP Bench. WP Bench measures how many WP queries per second and higher tends to be better (varies considerably by architecture). PHP Bench performs a lot of computational and some database operations which are measured in seconds to complete. Lower PHP Bench is better.

Results Table

Discussion

Servebolt had the fastest PHP bench and WP Bench scores by a significant margin. Servebolt run a patched version of MariaDB whereas the default version performs AES_ENCRYPT() function slower. I debated changing the plugin before I was aware it was patchable, but considering it is something fixable, I opted not to change it this year.

Nexcess had an incredibly fast WP Bench, the fastest of any WP Bench this year for a company not named Servebolt.

SSL Testing Results

The tool is available at https://www.ssllabs.com/ssltest/

Results Table

Discussion

Everyone got an A with Servebolt earning two A+'s.

Conclusion

There are two levels of recognition awarded to companies that participate in the tests. There is no ‘best’ declared, it’s simply tiered, it’s hard to come up with an objective ranking system because of the complex nature of hosting. These tests also don’t take into account outside factors such as reviews, support, and features. It is simply testing performance as described in the methodology.

Top Tier

This year's Top Tier WordPress Hosting Performance Award goes to the following companies who showed virtually no signs of struggle during the testing.

Honorable Mention

The following companies earned Honorable Mention status because they did very well and had a minor issue or two holding them back from earning Top Tier status.

No company achieved this status.

Individual Host Analysis

The Companies and Products

LoadStorm Testing Results

The Load Storm test (in k6) is designed to simulate real users visiting the site, logging in and browsing. It tests uncached performance.

The key metrics are:

• Total Requests - Number of requests k6 made
• Total Errors - Number of error requests
• p95 - The 95th percentile response time
• Page Cum Avg - Cumulative average response time for Pages (html)
• Asset Cum Avg - Cumulative average response time for Assets (css, js, images)
• Login Cum Avg - Cumulative average response time for Login (wp-profile.php)
• WP-Login Avg Response Time - Average response time for login (301 redirect which validates credentials)

Results Table

Discussion

GreenGeeks, Nexcess, Pressable, Presslabs, Seravo, Servebolt, SiteGround, and WordPress.com didn't have any issues.

Blallo started to show signs of load around 850 users and response times slowed down. It kept delivering, but it slowed down enough to keep it out of earning recognition.

Cloudways started slowing down around 350 VUsers with errors kicking in around 1,000.

Load Storm Average Response Time by Profile

This takes deeper look at the results and analyzes the performance of loading HTML mime types, to ignore a lot of static assets which are cached and bring down average response times. This tests how fast the initial pages are delivered, which is what a user would experience before loading all the other assets like css, javascript and images.

Results Table

Cloudways stands out with the highest response times.

Blallo had the second highest response times across every profile.

Presslabs had 2/4 of the fastest profiles along with fastest cart speed.

Pressable and Seravo picked up the two other fastest profile scores.

Servebolt was the second fastest in 2/4 profiles and 2nd fastest cart speed.

Nexcess and Raidboxes also slowed down a significant amount on every profile.

Pressable had 3/4 fastest profiles.

Servebolt was consistent top 3 in every profile.

ICDSoft was the fastest in one profile and top 3 in another.

WordPress.com was in the top 3 twice.

GreenGeeks had the fastest cart response time.

K6 Static Testing Results

K6 Static test is designed to test cached performance by repeatedly requesting the homepage.

Results Table

Discussion

Every single company handled the static test well. The entire cohort saw a combined 5 errors total which is astonishingly low. The highest average response time was 130ms which is still quite fast. Pressable and WordPress.com both were single digit average response times with 5 other companies following with 30ms or lower right behind them (Nexcess, Presslabs, Servebolt, SiteGround, Blallo).

Uptime Testing Results

Uptime is monitored by two companies: HetrixTools and Uptime Robot. A self hosted monitor was also run in case there was a major discrepancy between the two third party monitors.

Results Table

Discussion

Uptime Robot showed 100% uptime on every single company. HetrixTools showed above 99.9% on every company. Not a single company had uptime issues during this test.

WebPageTest Testing Results

WebPageTest fully loads the homepage and records how long it takes from 12 different locations around the world. Result are measured in seconds.

Results Table

Discussion

WordPress.com was the fastest this year on average but only was the fastest in Mumbai.

Presslabs earned itself second fastest average with 3/12 locations showing it to be the fastest.

Last year's fastest GreenGeeks, improved their average from 773ms to 730ms, a 43ms improvement. It was the fastest in 4/12 locations. The field continued to get faster and it only earned third this year on average response time.

SiteGround had the fastest in Sydney, Servebolt in Brazil, Blallo in Dubai and Cape Town.


Everyone, especially consumers, looks like a winner given the historical context.

WPPerformanceTester Testing Results

WPPerformanceTester performs two benchmarks. One is a WordPress (WP Bench) and the other is a PHP Bench. WP Bench measures how many WP queries per second and higher tends to be better (varies considerably by architecture). PHP Bench performs a lot of computational and some database operations which are measured in seconds to complete. Lower PHP Bench is better.

Results Table

Discussion

Servebolt had the fastest PHP bench and WP Bench scores by a significant margin. Servebolt run a patched version of MariaDB whereas the default version performs AES_ENCRYPT() function slower. I debated changing the plugin before I was aware it was patchable, but considering it is something fixable, I opted not to change it this year.

Presslabs and then Nexcess had the second and third fastest PHP bench scores. Then the field becomes pretty close with Blallo in the rear.

The WP Bench scores were much more distributed with WordPress.com and GreenGeeks taking second and third respectively.

Seravo had the slowest WP Bench but operates in a cluster, which is a good reminder that WPPerformanceTester is single threaded. Which is why it's a nice piece of information but doesn't necessarily determine real world performance.

SSL Testing Results

The tool is available at https://www.ssllabs.com/ssltest/

Results Table