What Is Google Lighthouse? A Current Guide

Originally published as “What Is Google Lighthouse 3.0?” Updated August 2026.*

Google Lighthouse is an open-source, automated tool that provides insights into how developers and online marketers can improve the quality of web pages. It audits a page’s performance, accessibility, SEO and adherence to web development best practices.

Overview of Lighthouse Audits

Lighthouse can help identify technical issues that affect the experience of visiting and using a website. Its main report categories in Chrome DevTools are Performance, Accessibility, Best Practices and SEO. However, automated testing has limits, particularly for accessibility and SEO, so its results should be treated as a starting point rather than a complete website audit.

Understanding How to Use Lighthouse

Auditing a web page with Lighthouse is simple if you know how to use Chrome DevTools. Similar to using Google Page Speed Insights, simply browse to the page in Chrome, open DevTools, and select the Lighthouse panel. Choose your preferred device mode and audit categories, then click “Analyse page load” to generate your report.

Lighthouse reloads and analyses the page. After approximately 30 to 60 seconds, it displays an audit report with a score for each selected category.

The Performance score is a weighted calculation based on several laboratory performance metrics. It is not a direct statement that your website is faster than a particular percentage of all websites. Lighthouse converts each measured value to a score using scoring curves based on real website performance data from the HTTP Archive.

The remaining category scores generally reflect whether the page passed the audits within each category. Best Practices is an official Lighthouse category that checks aspects of code health, browser security and modern web-development practices.

An audit may sometimes be marked as not applicable or return an error when Lighthouse cannot run it reliably. This does not necessarily mean that the page failed the audit, so the accompanying explanation should be reviewed.

The Performance report currently focuses on five scored laboratory metrics:

• First Contentful Paint (FCP): Marks the time when the first text or image is painted.
• Speed Index: Shows how quickly the visible contents of a page are populated.
• Largest Contentful Paint (LCP): Measures when the largest visible content element is rendered.
• Total Blocking Time (TBT): Measures how long the main thread was blocked during page loading and unable to respond promptly to user input.
• Cumulative Layout Shift (CLS): Measures unexpected movement of visible page content.

In the current scoring model, FCP and Speed Index each contribute 10% of the Performance score, LCP contributes 25%, TBT contributes 30%, and CLS contributes 25%. Lighthouse’s scoring model can change as the tool evolves, so the latest official documentation should be checked when exact weightings matter.

The report may also include a filmstrip showing step-by-step images of the page as it loads. This is useful for confirming that visible content loaded as expected. For more detailed investigation, developers can inspect the Lighthouse trace or use Chrome DevTools’ Performance and Network panels to analyse main-thread activity, requests and loading behaviour.

Following the performance review, Lighthouse provides opportunities, diagnostics and guidance for the selected categories. Some recommendations are technical, but the Learn more links explain why each audit matters and how the issue may be fixed.

The numerous automated checks are one reason Lighthouse is a valuable audit tool. It also highlights items that require manual checking, which are important reminders that no automated tool can fully assess a page’s quality, usability, accessibility or search performance.

What Does Lighthouse Bring to the Table?

Clearer Reports and Actionable Insights

Lighthouse reports use colour-coded scores and individual audit results to show which areas are performing well and which need attention. Scores from 90 to 100 are considered good, 50 to 89 need improvement, and 0 to 49 are considered poor.

Reports now provide more actionable insights into common problems such as render-blocking resources, inefficient image delivery, excessive JavaScript, layout instability and slow rendering of important content. The available audits and presentation can vary according to the Lighthouse version and the way the audit is run.

Performance Audit Scoring

A page receives a Performance score from 0 to 100. This score is a weighted average of the measured performance metrics, and the weightings have changed over time to better reflect users’ perception of performance.

Scores can vary between runs because of device performance, browser extensions, network conditions, advertising, A/B tests and other environmental factors. For this reason, it is better to run several comparable tests and treat performance as a range rather than relying on one isolated score. A perfect score of 100 is not required and should not be pursued at the expense of more meaningful improvements.

Additional and Updated Audits

Lighthouse continues to add, update and retire audits as web standards and browser capabilities evolve. It can identify opportunities involving images, JavaScript, CSS, security, accessibility, crawlability and other technical aspects of a page.

Some earlier Lighthouse 3.0 metrics—including First Meaningful Paint, First CPU Idle, Estimated Input Latency and Time to Interactive—have since been retired from the Performance score. Modern reports instead emphasise FCP, Speed Index, LCP, TBT and CLS.

Lighthouse and Core Web Vitals

Lighthouse is especially useful for diagnosing performance in a controlled laboratory environment, but it should be used alongside real-user field data. The current Core Web Vitals are:

• Largest Contentful Paint (LCP): 2.5 seconds or less.
• Interaction to Next Paint (INP): 200 milliseconds or less.
• Cumulative Layout Shift (CLS): 0.1 or less.

Core Web Vitals are assessed using real-user experiences at the 75th percentile, separately for mobile and desktop. PageSpeed Insights and Google Search Console can provide this field data when sufficient information is available. Lighthouse does not directly measure field INP during a page-load audit; Total Blocking Time is a laboratory metric that can help identify JavaScript work likely to affect responsiveness.

Faster and More Consistent Audits

Lighthouse can simulate mobile network and CPU conditions to estimate how a page may perform on a less powerful device. Although this makes testing faster and more repeatable, results can still vary. Tests should use consistent settings, and important changes should be verified with both repeat laboratory tests and real-user data.

The progressive improvements made to Lighthouse emphasise the importance of creating fast, accessible and reliable experiences on every device. Lighthouse makes it easier for developers and marketers to identify technical opportunities, but it does not replace a complete SEO, accessibility or user-experience review.

Good Lighthouse and Core Web Vitals results can support a better page experience, but they do not guarantee high Google rankings. Relevant, useful content and the website’s overall quality remain essential.