Docs7 vs. Mintlify: a speed benchmark

Docs7 won 33 of 34 Google PageSpeed comparisons, with 1 tie. DebugBear favored Docs7 on 9 of 9 mobile and 9 of 9 desktop page pairs.

I tested documentation with images, video, long guides, code examples, and custom components on both platforms. The charts show the results; full tables and report links are at the end.

Google ran 10 tests per URL and device. The chart above compares page medians. Below, each square compares the same page in the same numbered round. These are repeated tests of 17 page pairs; the two hosts were not always tested at the same time.

The size of the gap

Across the tested pages, Google's average score was 80.1 vs. 58.9 on mobile and 94.9 vs. 88.4 on desktop. Docs7 is listed first.

Google and DebugBear scores are out of 100; higher is better.

Page load times

Pingdom measures full page load time in milliseconds; lower is better. Docs7 loaded faster on 14 of 17 pages from San Francisco, 5 of 7 from Frankfurt, and 3 of 3 from Tokyo.

Each location covers its own set of page pairs. Bars show the average of the per-page medians.

Why these pages?

I chose pages to cover different types of documentation, then checked their content on both live sites before testing.

• A reference page with nine screenshots tests image loading and layout stability.
• A video page combines an embedded player with a custom component catalog.
• Four guides cover steps, release notes, a long CLI reference, and a setup flow with tabs and images.
• Five code and API pages cover long references, expandable fields, scheduling, retries, and SDK setup.
• Cards and a custom React badge test custom components. Separate examples cover tabs, accordions, and a file tree.
• A Mermaid page tests diagram rendering.

Results by page type

Each bar shows the average of the page scores in that content group.

Inside one result

These Google desktop results are from the video and component catalog page. Each value is the median of 10 tests.

How we tested

We compared 17 page pairs on September 17, 2026, using the direct Docs7 and Mintlify hosted URLs.

• Google PageSpeed Insights ran 10 tests per URL on mobile and desktop through its API, for 680 tests in total. All used Lighthouse 13.4.1. We kept every valid run. We use each page's median score for both hosts, following Google's guidance on repeated tests and medians.
• Pingdom tested 17 page pairs from San Francisco, 7 from Frankfurt, and 3 from Tokyo. San Francisco has three runs per URL for one pair, two for three pairs, and one for the rest. Frankfurt and Tokyo have one run per URL.
• DebugBear tested 9 mobile and 9 desktop page pairs from US East. Three mobile and two desktop pairs have two runs per URL; the rest have one.

We combine valid runs for the same URL, device, and location using the median. With two runs, the median is their average. Both hosts have the same run count in each comparison. Charts average the per-page results with equal weight for each page. Services and locations stay separate.

These lab tests measure initial page loads. Scores can vary between runs.

Full results

Page names link to Docs7. Small numbered links beside DebugBear and Pingdom values open each run's report.

Each cell is Docs7 / Mintlify, with the winner in bold.

Google PageSpeed Insights

Median scores out of 100, from 10 runs per URL and device. Download all 680 Google measurements as CSV.

DebugBear

Median performance scores out of 100, tested from US East.

Mobile

Desktop

Pingdom

Median load time in milliseconds. Lower is better.

San Francisco

Frankfurt

Tokyo