The State of Web Vitals
Real-user Core Web Vitals from 200,000 websites. Chrome field data, not lab tests — what visitors actually experience.
About 200,000 sites with complete Chrome field data in this edition. Start in the explorer, or pull the same numbers through MCP. How we measure is documented below.
Start here browser UI and agent API on the same dataset
Data explorer
Live charts for every Core Web Vital and technique metric we track — images, scripts, stack, network, headers, and more. Filter by CMS, framework, CDN, or third party.
Open the explorer →MCP endpoint
Streamable HTTP MCP. No auth — public research data. Wire it into Cursor, Claude, Grok, or any MCP client.
https://www.corewebvitals.io/api/state-of-cwv/mcp {
"state-of-cwv": {
"type": "http",
"url": "https://www.corewebvitals.io/api/state-of-cwv/mcp"
} Inside the explorer what you can open and compare
Core Web Vitals
LCP, INP, CLS, FCP, TTFB — pass rates, p75s, and full distributions for all sites or one tech slice.
Images, scripts & delivery
Loading modes, formats, origins, compression, cache strategy — how pages are built and shipped.
Stack & third parties
Frameworks, analytics, tags, consent, optimization plugins — what sites load and how common each is.
Cohorts
Scope any chart to WordPress, React, Cloudflare, GA4, and more. Same filters the MCP exposes.
MCP tools filter + group_by for agents
get_metrics— CWV, pass rates, technique paths. Optionalfilterandgroup_byget_histogram— distribution buckets for one vitallist_options— discover keys and metric paths
Per-site live RUM is CoreDash. This MCP is the State of CWV cross-site dataset.
How we measured field data, crawl, and this edition
Methodology
Chrome UX Report field metrics, what we crawl on each origin, thresholds, and how pass rates are defined.
This edition
Q2 2026 · about 200,000 sites with complete Chrome field data
Explorer and MCP read the same live snapshot — not a separate export.