Ultimate Website Screenshot, PDF, GIF and Video Capture

Capture any web page as a pixel-perfect JPEG, PNG, PDF document, animated GIF, or MP4 video — all from a single Apify actor, with 31 real device presets out of the box.

What this actor does

Feed this actor one or more URLs and it loads each page in a real browser at the device and viewport you choose, then produces the capture in the format you need. Assets are stored in the run's key-value store and referenced in the dataset via a direct download link — ready to embed in reports, pipe into a CDN, or attach to a ticket.

It supports full-page captures that stitch the entire scrollable height (for images and PDFs), animated captures that scroll the page across frames (for GIF/MP4), and 31 device presets covering everything from a 4K desktop monitor to an iPhone 15 Pro Max and a foldable Surface Duo. A custom viewport is also available if none of the presets match. Anti-bot resilience is built in: when a target page responds with a block/challenge page, the actor's auto proxy fallback detects it and silently retries the capture through a residential proxy session, so you get a real screenshot instead of a CAPTCHA image.

Everything is configurable — JPEG quality, PDF paper size and margins, navigation wait strategy, post-load delay, scroll-to-bottom behavior, cookie injection for session-walled pages, and MP4/GIF frame timing. The result is one consistent tool that replaces half a dozen specialized screenshot services.

Key features

• Five output formats — JPEG, PNG, PDF, GIF, MP4.
• 31 device presets — desktop monitors (1080p, 1440p, 4K), MacBook Pro 14/16, iPhone 13/14/15 families, Pixel phones, Galaxy phones, iPad Pro/Mini, Surface Duo, plus a Custom viewport option.
• Full-page capture that stitches the entire scrollable height for images and PDFs.
• Animated captures — GIF and MP4 with configurable frame count, interval, FPS, and scroll step per frame.
• Scroll-to-bottom trigger to force lazy-loaded images and content to render before capture.
• Auto proxy fallback — detects block/challenge pages and silently retries through a residential proxy so the run still produces a usable asset.
• Cookie injection for pages behind simple session walls.
• Proxy support (Apify Proxy or custom proxy URLs) for geo-specific or blocked pages.
• Configurable waits — load, domcontentloaded, or networkidle, plus a user-defined delay.
• PDF controls — A4 / Letter / Legal paper size, millimeter margins, background printing toggle.
• Graceful degradation — if MP4 encoding fails at runtime the actor falls back to a JPEG of the target page so your dataset is never empty.

Input

Field	Description
linkUrls	One or more http/https URLs to capture.
outputFormat	jpeg, png, pdf, gif, or mp4.
device	Any of the 31 presets, or Custom to use your own viewport.
windowWidth / windowHeight	Viewport size when device is Custom.
fullPage	Capture the entire scrollable page (applies to images + PDF).
waitUntil	Navigation wait strategy: load, domcontentloaded, or networkidle.
timeoutSeconds	Navigation timeout in seconds (1–120).
delayBeforeScreenshotMs	Extra wait after page load, in milliseconds (0–30000).
scrollToBottom	Scroll the page before capture to trigger lazy content.
jpegQuality	JPEG quality (1–100).
pdfFormat / pdfPrintBackground / pdfMargin*	Paper size, background toggle, margins (mm).
videoFrameCount, videoFrameIntervalMs, videoFps, videoScrollStepPx	Control GIF/MP4 length, timing, and scroll-per-frame.
cookies	Array of {name, value, domain?} cookies injected into the browser.
enableSSL	When false (default), invalid HTTPS certificates are ignored.
autoProxyFallback	When true (default), detect block/challenge pages and retry via residential proxy.
proxyConfiguration	Apify Proxy or custom proxy URLs.

Example input

{
  "linkUrls": [
    { "url": "https://news.ycombinator.com" },
    { "url": "https://example.com/pricing" }
  ],
  "outputFormat": "png",
  "device": "Desktop 1920x1080",
  "fullPage": true,
  "scrollToBottom": true,
  "waitUntil": "networkidle",
  "delayBeforeScreenshotMs": 1500,
  "autoProxyFallback": true
}

Output

Each capture produces one dataset record:

{
  "linkUrl": "https://example.com/pricing",
  "screenshotUrl": "https://api.apify.com/v2/key-value-stores/.../records/example-pricing.png",
  "storageKey": "example-pricing.png",
  "contentType": "image/png",
  "fileSizeBytes": 428913,
  "device": "Desktop 1920x1080",
  "viewportWidth": 1920,
  "viewportHeight": 1080,
  "format": "png",
  "capturedAtUTC": "2026-04-24T10:15:00+00:00"
}

Field descriptions

• linkUrl — the URL that was captured.
• screenshotUrl — direct download link to the captured asset in the run's key-value store.
• storageKey — key inside the run's key-value store (useful when referencing assets by convention).
• contentType — MIME type of the asset (image/jpeg, image/png, application/pdf, image/gif, video/mp4).
• fileSizeBytes — size of the stored asset in bytes.
• device — the device preset used (or Custom).
• viewportWidth / viewportHeight — actual viewport dimensions used for the capture.
• format — final format saved. May differ from the requested format if the MP4 fallback engaged.
• capturedAtUTC — ISO 8601 UTC timestamp of capture.

If a URL cannot be captured at all, a compact error record is written in its place with type: screenshot_error and a human-readable message, so your dataset is never silently empty.

Use cases

• Automated QA & visual regression — screenshot every critical page in your marketing site across desktop + mobile presets on every deploy.
• Archiving & compliance — snapshot competitor pricing pages, regulated content, or legal notices as dated PDFs for record keeping.
• Social media & preview cards — generate high-DPI screenshots of pages to embed in blog posts, Slack threads, and Notion docs.
• Marketing demos — record short MP4/GIF walkthroughs of a web app scrolling through its home page for pitch decks and ads.
• Accessibility & cross-device review — compare how the same page renders across iPhone 15 Pro Max, Pixel 8, and a 4K desktop without owning the hardware.

FAQ

Can it capture sites that block bots? Yes in most cases. The actor uses a real browser with a realistic user agent and respects the waitUntil and delayBeforeScreenshotMs settings so content renders before capture. When a target page responds with a block/challenge page, autoProxyFallback (on by default) retries the capture through a residential proxy session automatically.

How do I capture pages behind a login? Paste your session cookies as an array of {name, value, domain?} objects into the cookies field. If domain is omitted, the page's host is used.

Does full-page capture work for every format? Full-page applies to image (JPEG/PNG) and PDF formats. GIF and MP4 capture the current viewport across a sequence of frames while scrolling, so they reflect real on-screen animation.

What if MP4 encoding fails at runtime? The actor falls back to a JPEG of the target page and marks format: "jpeg" in the dataset record so your run still produces usable output rather than an error.

What's the difference between load, domcontentloaded, and networkidle? domcontentloaded is fastest (fires once the HTML is parsed), load waits for images and subresources, and networkidle waits until there are no network requests for 500ms — best for JS-heavy apps but slowest.

Will it capture every pixel of an infinite-scroll page? No tool can — infinite scroll has no end. Use scrollToBottom to trigger lazy-loading and set a reasonable page height cap via videoFrameCount (for MP4/GIF) to control how far the capture scrolls.

Do I pay per screenshot? Pricing is set in the Apify UI, separate from the actor itself. Costs depend on run duration, compute, and whether you use residential proxy.

Known limitations

• Infinite-scroll feeds (Twitter, TikTok) cannot be fully captured — full-page stitches only the currently-loaded DOM, not content that requires scrolling to materialize indefinitely.
• DRM-protected video (Netflix, DRM-locked streaming) cannot be recorded in MP4 captures — the browser blacks out protected frames.
• CAPTCHA gates that block all browser traffic (including real users) cannot be bypassed; those pages will be captured as-is or emit an error record.
• Fixed-position headers/footers can appear duplicated in full-page stitched screenshots on some sites — this is a common browser rendering limitation across all screenshot tools.
• Large MP4 captures (high frame count at high resolution) produce large files. Use reasonable videoFrameCount and videoFps values to keep storage manageable.