A plugin to run Lighthouse against a set of urls to see which site is the fastest.
npm install performance-leaderboard
- Median Run Selection:
performance-leaderboard
will run Lighthouse on the same site multiple times and select the Median run. It factors in First Contentful Paint, Largest Contentful Paint, and Time to Interactive when selecting the median run.
- Create a test file, say
sample.js
:
const PerfLeaderboard = require("performance-leaderboard");
(async function() {
let urls = [
"https://www.gatsbyjs.org/",
"https://nextjs.org/",
"https://www.11ty.dev/",
"https://vuejs.org/",
"https://reactjs.org/",
"https://jekyllrb.com/",
"https://nuxtjs.org/",
"https://gohugo.io/",
];
// Create the options object (not required)
const options = {
axePuppeteerTimeout: 30000, // 30 seconds
writeLogs: true, // Store audit data
logDirectory: '.log', // Default audit data files stored at `.log`
readFromLogDirectory: false, // Skip tests with existing logs
// onlyCategories: ["performance", "accessibility"],
chromeFlags: ['--headless'],
freshChrome: "site", // or "run"
launchOptions: {}, // Puppeteer launch options
}
// Run each site 3 times with default options
console.log( await PerfLeaderboard(urls) );
// Or run each site 5 times with default options
console.log( await PerfLeaderboard(urls, 5) );
// Or run each site 5 times with custom options
console.log( await PerfLeaderboard(urls, 5, options) );
})();
- Run
node sample.js
.
Sample Output
[
{
url: 'https://www.11ty.dev/',
requestedUrl: 'https://www.11ty.dev/',
timestamp: 1623525988492,
ranks: { hundos: 1, performance: 1, accessibility: 1, cumulative: 1 },
lighthouse: {
version: '8.0.0',
performance: 1,
accessibility: 1,
bestPractices: 1,
seo: 1,
total: 400
},
firstContentfulPaint: 1152.3029999999999,
firstMeaningfulPaint: 1152.3029999999999,
speedIndex: 1152.3029999999999,
largestContentfulPaint: 1152.3029999999999,
totalBlockingTime: 36,
cumulativeLayoutShift: 0.02153049045138889,
timeToInteractive: 1238.3029999999999,
maxPotentialFirstInputDelay: 97,
timeToFirstByte: 54.63900000000001,
weight: {
summary: '14 requests • 178 KiB',
total: 182145,
image: 124327,
imageCount: 10,
script: 7824,
scriptCount: 1,
document: 30431,
font: 15649,
fontCount: 1,
stylesheet: 3914,
stylesheetCount: 1,
thirdParty: 15649,
thirdPartyCount: 1
},
run: { number: 2, total: 3 },
axe: { passes: 850, violations: 0 },
}
]
In the return object you’ll see a ranks
object listing how this site compares to the other sites in the set. There are a bunch of different scoring algorithms you can choose from:
ranks.performance
- The highest Lighthouse performance score.
- Tiebreaker given to the lower SpeedIndex score.
ranks.accessibility
- The highest Lighthouse accessibility score.
- Tiebreaker given to lower Axe violations.
- Second tiebreaker given to highest Axe passes (warning: each instance of an Axe rule passing is treated separately so this will weigh heavily in favor of larger pages)
ranks.hundos
- The sum of all four Lighthouse scores.
- Tiebreaker given to the lower Speed Index / Total Page Weight ratio.
ranks.cumulative
(the same ashundos
but with an Axe tiebreaker)- The sum of all four Lighthouse scores.
- Tiebreaker given to the lower Axe violations.
- Second tiebreaker given to the lower Speed Index / Total Page Weight ratio.
v4.0.0
Major version upgrade oflighthouse
dependency from v6.5 to v7.2v4.1.0
Updatelighthouse
to v7.3v5.0.0
Updatelighthouse
to v8.0v5.1.0
AddsaxePuppeteerTimeout
option. AddscarbonAudit
option.v5.2.0
Updatelighthouse
from v8.0 to v8.2v5.3.0
Updatelighthouse
from v8.2 to v8.5v9.0.0
Updatelighthouse
to v9.0. RemovescarbonAudit
, upstream API was removed.v9.2.0
Updatelighthouse
to v9.2.