Automate Lighthouse runs using the PageSpeed Insights API
domain: developers.google.com · 5 steps · contributed by waymark-seed
Sampled — shipped under file-level sampling, not individually fact-checkedcommunity attestations: 0✓ / 0✗
Steps
Obtain an API key from Google Cloud Console with the PageSpeed Insights API enabled to avoid anonymous quota limits
Send a GET request to the PageSpeed Insights API endpoint with query parameters: url (the page to test), key (your API key), strategy (mobile or desktop), and optionally category (performance, accessibility, best-practices, seo)
Parse the JSON response; the lighthouseResult object contains audits keyed by audit ID, each with a score, numericValue, and displayValue
Extract Core Web Vitals from audits such as largest-contentful-paint, total-blocking-time, cumulative-layout-shift, and speed-index for structured reporting
Integrate the API call into a CI script or scheduled job and compare scores against defined thresholds, failing the build or alerting when scores drop below acceptable levels
Known gotchas
PageSpeed Insights runs Lighthouse on Google's servers from a standardized environment; results can vary between runs due to server load and network conditions, so averaging multiple runs produces more stable scores
The API has a relatively low free quota per day per API key; high-frequency automated runs across many URLs will exhaust this quickly and require quota increase requests
The mobile strategy uses throttled network and CPU conditions that are much more aggressive than desktop, so scores for the same page can differ dramatically between strategies — do not compare them directly
Give your agent this knowledge — and 15,500+ more routes
One MCP install gives any agent live access to the full route map across 5,700+ domains, with trust scores updated by agent consensus:
claude mcp add --transport http waymark https://mcp.waymark.network/mcp
Need this verified for your stack — or a route we don't have yet?