Website Screenshot API — Full Page, Mobile, PDF, No Login

Take a website screenshot of any list of URLs in headless Chromium and get the image back with proof attached: a public image URL, final URL, HTTP status, page title, capture timestamp, browser build and SHA-256 hash. Full-page or first screen; desktop, laptop, tablet or phone presets; PNG, JPEG, WebP or PDF; dark mode; hide cookie banners, chat widgets and ads; wait for a selector, delay, scroll to load lazy content; capture one element or a clip. Pay per capture — blocked or unreachable pages are reported and never billed.

• Batch: up to 100 URLs per run, one row and one stored image each.
• Every capture is evidence: timestamp, final URL after redirects, HTTP status, browser version and SHA-256 on the row.
• No login, no CAPTCHA solving: each site's robots.txt is checked first; the browser identifies itself with a contact User-Agent. Optional Apify Proxy for sites that block data-centre IPs.
• Direct image links: image_url opens the stored PNG/JPEG/WebP/PDF straight from the run's key-value store.

Billing: $0.02 per run start, then $0.05 per completed capture (screenshot_check).

Who uses this

• Compliance, legal and brand teams that need dated website screenshots as evidence.
• QA, SEO and monitoring teams screenshotting page lists on a schedule (mobile and desktop, light and dark).
• Agencies and researchers archiving how pages looked, or turning pages into PDFs.

📊 Sample Output

Real rows from run cUFIUzrvtGqMjlhdw on build 0.1.36 (2026-10-18), the Quick start input below:

target_url	outcome	http_status	page_title	device	format	width×height	image_sha256
https://example.com/	answer	200	Example Domain	desktop	png	1280×800	018344bad235b72b…
https://www.iana.org/	answer	200	Internet Assigned Numbers Authority	desktop	png	1280×800	6f08a23db67efe72…

The run finished with the status message: 2 URL(s): 2 capture(s) stored+charged, 0 unbilled (blocked/unreachable/unparseable) | screenshot_check charged=2

✅ What you get

One screenshot_evidence row per URL (every URL gets a row; only outcome: answer rows are billed):

• target_url, final_url, http_status, page_title — what was asked for, where it landed, and what the page called itself.
• image_url, image_key, image_bytes, image_sha256 — the stored capture (public URL in this run's key-value store) and its hash.
• format (png/jpeg/webp/pdf), device, full_page, width, height, device_scale_factor — how it was captured.
• capture_timestamp, browser, browser_version, load_time_ms — when and with what.
• outcome — answer (captured, billed), blocked (robots / 403 / 429 / 5xx / private host — not billed), unreachable (DNS, timeout, navigation or capture fault — not billed), unparseable (not a usable URL — not billed), with outcome_reason.
• render_warnings — e.g. a waitForSelector that never appeared, or an output truncated at the size cap.

⚙️ Sample inputs

1. Quick start — the Store example

{
  "urls": ["https://example.com/", "https://www.iana.org/"],
  "device": "desktop",
  "blockCookieBanners": true,
  "maxUrls": 10
}

About $0.12 on the Free tier ($0.02 start + 2 × $0.05).

2. Full-page mobile capture with lazy content loaded, as JPEG

{
  "urls": ["https://www.iana.org/"],
  "device": "iphone",
  "fullPage": true,
  "scrollToBottom": true,
  "format": "jpeg",
  "quality": 75
}

3. Page to PDF, dark mode, German locale

{
  "urls": ["https://example.com/"],
  "format": "pdf",
  "colorScheme": "dark",
  "locale": "de-DE",
  "timezoneId": "Europe/Berlin"
}

4. One element only, after the app has settled, with widgets hidden

{
  "urls": ["https://example.com/pricing"],
  "waitUntil": "networkidle",
  "waitForSelector": "#pricing",
  "selector": "#pricing",
  "hideSelectors": ["#intercom-container", ".newsletter-popup"],
  "deviceScaleFactor": 2
}

5. Retina desktop-HD capture through Apify residential proxy

{
  "urls": ["https://example.com/"],
  "device": "desktop-hd",
  "deviceScaleFactor": 2,
  "blockAds": true,
  "proxyConfiguration": {"useApifyProxy": true, "apifyProxyGroups": ["RESIDENTIAL"]}
}

Older inputs still work: viewport: "desktop" | "mobile" | "full".

🧾 JSON sample record

{
  "target_url": "https://example.com/",
  "final_url": "https://example.com/",
  "http_status": 200,
  "page_title": "Example Domain",
  "viewport": "desktop",
  "device": "desktop",
  "format": "png",
  "full_page": false,
  "device_scale_factor": 1,
  "load_time_ms": 82,
  "capture_timestamp": "2026-10-18T22:04:25.809522+00:00",
  "browser": "chromium",
  "browser_version": "156.0.8078.4",
  "image_key": "screenshot-0-example.com.png",
  "image_url": "https://api.apify.com/v2/key-value-stores/UFzFi4JAeoyUMYV3S/records/screenshot-0-example.com.png",
  "image_bytes": 71152,
  "image_sha256": "018344bad235b72b58d300df236525ae98ba51b567899a10d5dbcad52e10fba5",
  "width": 1280,
  "height": 800,
  "render_warnings": null,
  "faults": null,
  "record_type": "screenshot_evidence",
  "source": "buyer-submitted page URLs",
  "outcome": "answer",
  "input": "https://example.com/",
  "outcome_reason": null
}

🔧 How it works

Capture. Headless Chromium (Playwright) opens each URL in a fresh context with the chosen device size, pixel density, colour scheme, locale and timezone; waits for waitUntil (+ optional selector, scroll pass and delay); applies hide/custom CSS; then writes PNG/JPEG (Chromium), WebP (converted from a lossless PNG) or PDF (print to PDF with backgrounds). Captures above 12 MB are truncated with a warning — lower the size, scale or use JPEG.

Gates before the browser opens a URL. The host must resolve only to public addresses (no localhost, private ranges or cloud metadata), and the site's robots.txt must allow the path for our user agent. A disallowed or non-public target is a blocked row, never a capture.

Blocking. blockAds aborts requests to well-known ad/tracking hosts; blockImages skips image and media downloads; blockCookieBanners hides common consent containers by CSS — nothing is clicked, accepted or bypassed.

Proxy. Direct by default. proxyConfiguration turns on Apify Proxy (datacenter or residential groups, optional country) for sites that refuse data-centre IPs; no third-party proxies.

Billing. Each completed capture is pushed and charged in one call; a row that cannot be charged is not delivered. Blocked, unreachable and unparseable rows are delivered unbilled so you always see every URL's disposition.

Limits. 100 URLs per run; navigation timeout 3-90 s per page; a page that fails to load or capture is unreachable (not billed) and the run continues.

💰 Pricing example

Event	Free	Bronze	Silver	Gold
Actor Start (apify-actor-start)	$0.02	$0.02	$0.02	$0.02
Completed capture (screenshot_check)	$0.05	$0.045	$0.04	$0.0335

• 2 captures: $0.02 + 2 × $0.05 = $0.12
• 100 captures: $0.02 + 100 × $0.05 = $5.02

Blocked and unreachable URLs cost nothing beyond the start fee.

⚖️ Legal & ToS

Captures public pages exactly as any visitor's browser would render them, checks each site's robots.txt first, identifies itself with a contact User-Agent and never logs in, solves CAPTCHAs or bypasses paywalls. Use the captures in line with each site's terms and your local law.

❓ FAQ

Q: Where is the image? A: image_url on each row is a direct link to the stored file in the run's key-value store (image_key is its name there). Files stay as long as the run's storage does (7 days on the Free plan, longer on paid plans) — download them if you need them permanently.

Q: Full page came out short / missing images. A: Turn on scrollToBottom (lazy-loaded sections render as the page scrolls) and, for single-page apps, waitUntil: "networkidle" plus a delayMs.

Q: The site shows a bot wall or 403. A: That URL is returned as blocked and not billed. Try proxyConfiguration with Apify residential proxy; the actor never solves CAPTCHAs.

Q: Can I capture only part of a page? A: Yes — selector for one element, or clip for a fixed rectangle.

Q: PDF? A: format: "pdf" prints the page with backgrounds; image_url then points to the PDF.

Q: What formats can I export? A: Rows download as JSON, CSV, Excel or XML from the Dataset tab or API; images via their image_url.

🆘 Troubleshooting

• outcome = unreachable, "navigation failed: TimeoutError" → raise timeoutMs or use waitUntil: "domcontentloaded".
• outcome = unreachable, "selector matched nothing" → check the CSS selector; use waitForSelector for late-rendered elements.
• outcome = blocked, robots → the site disallows that path for crawlers; the actor will not capture it.
• Run TIMED-OUT → many slow pages: lower maxUrls, shorten timeoutMs, or raise the run timeout.

🔗 Related Actors

• HTML to PDF Evidence Renderer — render your own HTML to PDF with a hash.
• Responsive Design Testing — one verdict per URL across desktop, tablet and mobile.
• Wayback Machine Website History — how a page looked in the past.
• 🏢 About NexGen Watch — more than 300 public-data actors, pay-per-result: apify.com/nexgenwatch.

⭐ Found this useful?

A quick review on the Apify Store helps other teams find it. Feature request? Open it from the Issues tab — every one is read.