Lighthouse

Visit Website

Introduction

Lighthouse is Google’s open-source, automated tool for auditing web pages. The Chrome for Developers documentation describes audit categories for performance, accessibility, best practices and SEO, and the project lives at GoogleChrome/lighthouse on GitHub under the Apache-2.0 licence, with about 31,000 stars. The latest release at review was v13.6.0 (27 October 2026), and the repository had commits in November 2026.

You can run Lighthouse in several ways: from the Lighthouse panel in Chrome DevTools, from the command line with the lighthouse npm package, as a Node module in your own scripts, or through PageSpeed Insights, which uses Lighthouse for its lab data. Each run loads the page in a controlled environment, with simulated throttling by default, and produces a report with scores from 0 to 100 per category, metric values such as Largest Contentful Paint, Total Blocking Time and Cumulative Layout Shift, and a prioritised list of opportunities and diagnostics.

Accessibility audits are powered by axe-core rules, so Lighthouse doubles as a quick accessibility smoke check, although it only catches issues that can be detected automatically. Best-practices audits flag problems such as console errors, deprecated APIs and insecure requests, and SEO audits check basics such as meta descriptions, crawlability and link text.

How it compares: Lighthouse is a lab tool, so its numbers vary between runs and machines. PageSpeed Insights adds Chrome UX Report field data, WebPageTest gives deeper waterfalls and real-browser testing from many locations, and sitespeed.io and Lighthouse CI add repeatable runs, budgets and dashboards. Many teams use Lighthouse locally during development and Lighthouse CI or a monitoring tool to catch regressions.

In the quality loop Lighthouse fits the performance and accessibility stages, and with Lighthouse CI it can act as a release gate.

Features

1. Four audit categories: Performance, accessibility, best practices and SEO.
2. Multiple entry points: Chrome DevTools, CLI, Node module and PageSpeed Insights.
3. Core Web Vitals lab metrics: LCP, TBT, CLS and other metrics with diagnostics.
4. Opportunities and diagnostics: Prioritised suggestions with estimated savings.
5. Throttling and device emulation: Mobile and desktop presets with network and CPU throttling.
6. Report formats: HTML, JSON and CSV output for automation.
7. User flows: Audit multi-step journeys, not just page loads.
8. Extensible: Custom audits and plug-ins via the Node API.

Pros

1. Free and open source: Apache-2.0 and built into Chrome.
2. Actively maintained: Frequent releases; v13.6.0 in October 2026.
3. Ubiquitous: The de facto standard for quick web audits.
4. Automation-friendly: JSON output and Node API suit CI.

Cons

1. Lab variance: Scores fluctuate between runs and environments.
2. Not field data: Does not reflect real users on its own.
3. Limited accessibility coverage: Automated checks catch only part of WCAG issues.

Usage and Links (Example)

Quick start: open Chrome DevTools, choose the Lighthouse panel, select categories and device, and generate a report. On the command line:

npm install -g lighthouse
lighthouse https://example.com --output html --output-path ./report.html

For CI, run several times and use the median, or adopt Lighthouse CI to store results and assert budgets on pull requests.

Useful links:

• Lighthouse documentation
• GitHub repository

Pricing

Free and open source under the Apache-2.0 licence, built into Chrome DevTools at no cost.

Recommended for?

• Front-end developers checking performance and accessibility during development
• Teams adding performance budgets to CI through Lighthouse CI
• SEO and marketing teams auditing page quality
• Anyone who needs a quick, free web page audit

Conclusion

Lighthouse is the standard free tool for auditing web performance, accessibility, best practices and SEO. Its reports are actionable and easy to automate, and it is maintained by the Chrome team.

Treat its scores as lab signals: combine Lighthouse with field data from PageSpeed Insights or RUM, and run it repeatedly in CI to spot trends rather than reacting to a single score.