Google PageSpeed Insights API
Browse all Google APIs✓ Official Vendor SpecAnalyticsWeb Analyticsoauth21 EndpointsREST
For Agents
Run a Lighthouse-based performance audit on any public URL and read back Core Web Vitals, lab metrics, and optimisation suggestions.
Use for: Audit the homepage performance for mobile users, Get the Largest Contentful Paint for a checkout page, Compare desktop and mobile PageSpeed scores for a URL, Find accessibility issues on a marketing landing page
Not supported: Does not run real-user monitoring, audit authenticated flows, or fix performance issues — use for synthetic Lighthouse audits of public URLs only.
The PageSpeed Insights API runs Lighthouse against any public URL and returns a structured performance report including Core Web Vitals, lab metrics, and optimisation suggestions. It exposes a single runPagespeed endpoint that accepts a URL plus options for strategy (mobile or desktop) and the categories to score (performance, accessibility, best-practices, SEO, PWA). Use it to integrate web performance budgets into CI, build dashboards that track real-user metrics over time, or pull suggestions to feed engineering backlogs. The spec exposes 1 endpoint.
Install Jentic One Beta
Connect the PageSpeed Insights API to your agent
Jentic One is a self-hosted execution layer for AI agents. It lets your agent call the PageSpeed Insights API, or any other public or private API you need. You set the rules, the agent never sees your credentials, and every call is logged.
Two steps, two machines. Install the instance in a safe environment, then register your agent from wherever it runs.
1
Step 1: Jentic One Host machine
# On the machine that will host your Jentic One instance:
curl -fsSL https://raw.githubusercontent.com/jentic/jentic-one/main/tools/install.sh | sh2
Step 2: Agent machine
# On the machine where your agent runs (keep this separate from the instance):
curl -fsSL https://raw.githubusercontent.com/jentic/jentic-one/main/tools/install.sh | sh
jentic register # connects your agent to your Jentic One instanceJentic One is in public beta. The setup above keeps your agent separate from the instance, which is what you want before using real credentials: an agent running as the same OS user as Jentic One can read its stored keys directly. Just evaluating? A single local install is fine to start. See the secure deployment guide for the tiers.
Capabilities
What an agent can do with PageSpeed Insights API.
Run a mobile or desktop Lighthouse audit against any public URL
Read Core Web Vitals (LCP, INP, CLS) along with lab and field metrics
Pull category scores for performance, accessibility, best-practices, SEO, and PWA
Surface specific audit failures with their affected DOM nodes and savings estimates
Capture screenshot timelines and resource summaries for a tested page
Use Cases
Patterns agents use PageSpeed Insights API for, with concrete tasks.
★ Performance budgets in CI
Run PageSpeed Insights against staging URLs after every deploy and fail the build if the performance score drops below a budget or if a Core Web Vital regresses. The single runPagespeed endpoint returns enough structured data — category scores plus per-audit results — to assert specific thresholds. Pairs naturally with PRs that compare current vs main scores.
Call GET /pagespeedonline/v5/runPagespeed with url=https://staging.example.com and strategy=MOBILE, parse the performance score, and fail the build if it is below 0.85.
Field metrics dashboards
Aggregate Core Web Vitals from PageSpeed Insights — including the CrUX field data — into a dashboard that tracks LCP, INP, and CLS for a portfolio of URLs over time. The structured response makes it cheap to extract metrics on a daily cron without scraping the public PSI page. Useful for content sites, e-commerce, and marketing teams reporting on page experience.
For every URL in homepages.json, call runPagespeed with strategy=MOBILE, extract loadingExperience metrics, and append them to a BigQuery table for trend reporting.
Optimisation suggestion feed for engineering
Pull the audits array from a PageSpeed Insights response to surface actionable recommendations — image format conversion, unused JavaScript, render-blocking resources — and turn them into engineering tickets. Each audit comes with details on the offending elements, byte savings estimates, and links to documentation.
Call runPagespeed for the URL https://example.com/products, extract any audit with score below 0.5 and weighted savings above 100ms, and create a Jira ticket for each.
Agent-driven page health checks via Jentic
A site-reliability agent reacting to a slow-page complaint can use Jentic to call PageSpeed Insights and reply with a precise breakdown of what regressed. Credentials stay in the Jentic vault, so the agent runs under a scoped key with audit trail. Useful for support bots that need to triage performance complaints automatically.
Use Jentic to find the run-pagespeed operation, load its schema, and execute it for the complained-about URL with strategy=MOBILE, then reply with LCP, INP, and CLS values.
Key Endpoints
1 endpoints — the pagespeed insights api runs lighthouse against any public url and returns a structured performance report including core web vitals, lab metrics, and optimisation suggestions.
METHOD
PATH
DESCRIPTION
GET
/pagespeedonline/v5/runPagespeed
Run a Lighthouse audit against a URL
GET
/pagespeedonline/v5/runPagespeed
Run a Lighthouse audit against a URL
Why Jentic?
Three things that make agents converge on Jentic-routed access.
Credential isolation
PageSpeed API keys are stored encrypted in the Jentic vault (MAXsystem). Agents call the API via Jentic's broker; the raw key never enters their context.
Intent-based discovery
Agents search Jentic with intents like audit page performance and Jentic returns the runPagespeed operation, its query parameters, and the response schema.
Time to first call
Direct integration: a few hours to wire the key, parse the response, and parse Lighthouse results. Through Jentic: under 15 minutes — search, load, execute.
Related APIs
Alternatives and complements available in the Jentic catalogue.
FAQs
Specific to using PageSpeed Insights API through Jentic.
What authentication does the PageSpeed Insights API use?
The API supports anonymous use with a basic quota and Google OAuth 2.0 / API key for higher quotas. Through Jentic, the API key is stored encrypted in the Jentic vault and attached to each request, so agents never see the raw key.
Can I get Core Web Vitals from this API?
Yes. The runPagespeed response includes a loadingExperience field with field-level Core Web Vitals (LCP, INP, CLS) sourced from the Chrome User Experience Report when enough field data is available.
What are the rate limits for the PageSpeed Insights API?
Anonymous calls are limited to a small number of queries per second; with an API key the default is 25,000 queries per day per project. For higher volume, request a quota increase in the Cloud Console rather than retrying on 429.
How do I run an audit through Jentic?
Search Jentic for run pagespeed audit, load the schema for GET /pagespeedonline/v5/runPagespeed, and execute it with the URL, strategy, and category list. Jentic returns the structured Lighthouse report for parsing.
Is the PageSpeed Insights API free?
Yes — the public PSI API is free within the documented quota. You only pay if you wrap it inside another billable Google Cloud product like Cloud Run for orchestration.
Does this API audit pages behind authentication?
No — runPagespeed only reaches public URLs. To audit authenticated flows, run Lighthouse locally with cookies or a service-worker, then use this API only for public benchmarks.
GET STARTED