How does this handle React and Vue rendered pages?

A real headless Chromium instance loads the URL, runs the JavaScript, waits for the page to settle, then we extract from the rendered DOM. Same result as if a user had opened the page in a browser — including content from useEffect hooks, dynamic imports, and lazy-loaded components.

Does this work on Next.js client-rendered routes?

Yes — pages using getServerSideProps or full SSR work fine via static fetch (we try that first, it's faster and cheaper). Pages using client-side data fetching (useSWR, useQuery, plain useEffect) need the headless path, which we use automatically when the static fetch returns an empty shell.

How long does JS rendering take vs static fetch?

Static fetch: ~200–500ms per page. Headless rendering: ~2–5 seconds per page (browser launch dominates the first request, then amortises). The MDisBetter web tool runs the headless path automatically when needed; if you're rolling your own at scale, Playwright reuses a single browser context across requests for the same speedup.

Can I configure how long to wait for content to load?

Yes — three strategies: networkidle (default), selector (wait for a CSS selector — ?wait_for=.product-grid), or timeout (fixed delay — ?wait_ms=3000). Selector-based waits are the most reliable; use them when you know the page has a recognisable "ready" element.

Will infinite-scroll pages capture all content?

By default no — we render the initial viewport. Pass ?scroll=true and we auto-scroll the page in the headless browser, triggering the lazy-load until the page stops growing. Useful for product catalogs and feed-style sites; expect higher latency proportional to total page length.

JavaScript SPA to Markdown — React, Vue, Angular

How we detect a JS-required page

Two signals. First, we attempt a static fetch and look at the response. An almost-empty body (<div id="root"></div>, <div id="__next"></div>, near-zero text content, a script tag fetching main.js) is the obvious tell. Second, we look for known SPA framework markers (data-reactroot, ng-version, __NUXT__) and route those URLs through the headless path automatically. You can also force browser rendering with ?render=js.

Wait strategies

The hardest part of SPA scraping isn't rendering — it's knowing when rendering is "done". We support three strategies: networkidle (wait until network requests stop, default), selector (wait until a specific CSS selector appears in the DOM — pass ?wait_for=h1), and timeout (fixed delay in ms, last resort for poorly-instrumented apps). Most pages settle within 2–4 seconds.

JavaScript SPA to Markdown — Convert Dynamic Pages

How we detect a JS-required page

Wait strategies

Before / After

Frequently asked questions

Stop feeding garbage
to your AI

Tools

Stop feeding garbageto your AI

Tools

Stop feeding garbage
to your AI