Website Screenshot API — Clean, No Popups

Take clean, complete screenshots of many websites in one run. Paste your links, press start, and get one image and one clear result row for every link. It is made for people who need good screenshots at scale: marketers, designers, QA testers, researchers, and anyone who keeps records of web pages.

Cookie banners, popups and chat bubbles are hidden. Full pages are really full. Every failure comes with a reason in plain words.

Why it is better

• Clean screenshots. Cookie banners, newsletter popups and chat bubbles are hidden. You can also ask for a click on "Reject all". The Actor never clicks "Accept all".
• Real full page. Tall pages are taken slice by slice and joined, so there are no repeated headers. Pages that scroll inside a box are unrolled too.
• Lazy images loaded. The page is scrolled slowly first, so pictures that load late are in the image.
• Endless page guard. Pages that never stop loading are detected. You choose: keep the top part, or skip the page.
• All languages and emojis. Hindi, Arabic, Chinese, Japanese, Korean, Thai, Hebrew and more, plus colour emojis.
• Clear reasons. If a page fails, the row says what happened, why, and what to do.
• One row per link, always. Even when a run is stopped, every link gets a row.
• Safe by design. The Actor never opens private or internal addresses, never types into pages, and never logs in.

Quick start

1. Paste your links in Website links (one per line, up to 500).
2. Pick Device: Laptop or Phone.
3. Pick Capture: Full page or Top only.
4. Press Start.

Field	What it does	Default
Website links	1 to 500 links. https:// is added if missing. Duplicates are removed.	https://example.com
Device	Laptop = 1440 x 900. Phone = 390 x 844, sharp (2x), mobile layout.	Laptop
Capture	Full page = top to bottom. Top only = the first screen.	Full page

Example input

{
    "urls": ["https://example.com", "https://www.wikipedia.org"],
    "device": "phone",
    "capture": "fullPage"
}

Example output row

{
    "url": "https://example.com/",
    "finalUrl": "https://example.com/",
    "result": "Done",
    "notes": "",
    "charged": false,
    "imageUrl": "https://api.apify.com/v2/key-value-stores/<store-id>/records/screenshot-94f871ba1cb95ff631b85415.png",
    "imageKey": "screenshot-94f871ba1cb95ff631b85415.png",
    "format": "png",
    "width": 780,
    "height": 1796,
    "device": "Phone",
    "takenAt": "2026-11-06T16:30:36.393Z",
    "errorType": null,
    "whatHappened": null,
    "why": null,
    "whatToDo": null
}

Advanced settings

All of these are optional. They sit in one collapsed section of the form.

Group	Field	Values	Default
Size and view	Device preset	None, iPhone, Android phone, iPad, Laptop, Big monitor (2560 x 1440). Overrides Device.	None
	Custom width / height	Width 320 to 3840, height 320 to 2160. Fill in both. Overrides the preset.	empty
	2x sharp mode	On or off. Only when the width is 1920 or less.	Off
Capture	Capture mode	Use main field, Visible screen, Full page, Top N px, One element (CSS selector). Overrides Capture.	Use main field
	Top N px	100 to 30000	3000
	Element CSS selector	Up to 300 characters	empty
Waiting	Max wait time	5 to 60 seconds	30
Cleaning	Clean screenshot	Hides popups, newsletter boxes and chat bubbles	On
	Cookie banner	Hide, or Click "Reject all"	Hide
	Sticky elements	Keep once at top, Hide, Leave as is	Keep once at top
Endless pages	When found	Capture top, or Skip	Capture top
	Max page height	1000 to 30000 px	15000
Pagination	Follow pages up to N	0 to 20	0
	Click "Load more" up to N times	0 to 20	0
Clicks	Run these clicks first	Up to 10 steps	empty
Output	Format	PNG, JPEG, WebP, PDF	PNG
	Resize output width	100 to 3840 px, keeps the shape	empty

Notes:

• Device preset uses ready-made device settings (iPhone, Pixel, iPad). Custom width and height beat the preset, and the preset beats Device.
• Sticky elements: "Keep once at top" shows a fixed header once at the top, not again and again down the image.
• Follow pages: the Actor follows the page's "Next" link (or a link or button with text like Next, >, >>, or the next page number). Same website only. Each page gets its own row, for example "Page 2 (followed Next ...)".
• Load more: presses buttons with text like Load more, Show more, See more or View more.
• Run these clicks first: each step is the exact text of a button or link. Capital letters and extra spaces don't matter. Experts can write css:<selector>. Each step has 5 seconds. Text is up to 100 characters and a selector up to 300. The Actor never types anything.
• Cookie banner → Click "Reject all": works in many languages. If no reject button is found, the banner is hidden instead and a note says so.

Output

Every link gets one row in the results table. The images are saved in the run's key-value store, and each row has a link to its image.

Column	Meaning
URL	The link, after clean-up
Result	The status (see below)
What we noticed	Notes, for example "Cookie banner hidden" or "Max wait hit"
Charged?	false for every row while the Actor is free
Image	Link to the picture

The All details view adds: final URL (after redirects), image key, format, width, height, device, time taken, and for problems: error type, what happened, why, and what to do.

Statuses

Status	Meaning	Charged?
Done	Full screenshot delivered	Counts as delivered
Cut short	Endless or too-tall page; the top part is delivered	Counts as delivered
Partial	Image saved, but something went wrong (a click step was not found, the max wait was hit, or similar)	No
Blocked	The site blocked us (for example 403), or the address is private	No
Failed	Any error from the table below	No
Skipped	Endless page with Skip, an age check, a login page, or the run stopped before this link	No
Invalid link	The link failed the check before the run	No

While the Actor is free, the Charged? column is false on every row, because nothing is charged. The run summary still counts the delivered screenshots (Done and Cut short).

Run summary

At the end of each run the Actor saves a SUMMARY record in the key-value store and shows it as the run status message. It has the number of links per status, delivered screenshots, time taken, and compute used.

Downloads

Open the Storage tab of the run (or use the API) and download the results table as JSON, CSV or Excel.

Keep your images

Apify keeps the storage of a run for a limited time. Download the images you need to keep.

Failure reasons

What happened	Why	What to do
Invalid link	The link is not a valid web address	Fix the link and run again
Domain doesn't exist	No website with this name	Check the spelling
404 Not found	The site works, the page does not exist	Check the link in your browser
403 Blocked	The site refuses automated tools	Try later; some sites never allow it
500 Server error	A problem on the site's side (we tried twice)	Try later
Timeout	The site took too long (we tried twice)	Try later, or raise Max wait time
SSL certificate problem	The security certificate is expired or wrong	Tell the site owner
Redirect loop	The page sent us on more than 10 times	Use the final link
Private address blocked	The link points to a private or internal address	Use a public website link
Too big	The page tried to download more than about 50 MB	Try Top only
Page crashed	The page used too much memory	Try Top only. If you lowered the memory, run with 2048 MB (the default)
Connection problem	The connection failed (we tried twice)	Try later
Element not found	Your CSS selector matches nothing visible	Check the selector
Endless page skipped	Endless page, and the setting is Skip	Set it to Capture top
Age check found, not clicked	The page asks for age; the Actor never clicks it	See "Age checks" below
Login needed	The page needs a login	Use a public page
Not finished in time	The run was stopped before this link was done	Run it again

Limits

Limit	Value
Links per run	500 (duplicates removed), up to 2048 characters each
Pages at once	3 overall, 2 per website, with a short pause between pages of the same site
Time per link	About 90 seconds
Screen size	Width 320 to 3840, height 320 to 2160
Page height	Up to 30000 px (default 15000)
Total image size	About 40 million pixels
Click steps	10 per link; Load more 20; Follow pages 20 (total pages per run 500)
Max wait time	5 to 60 seconds
Image file size	About 8.5 MB. Bigger images are made smaller and a note says so
Links allowed	http and https, ports 80 and 443 only
Memory	2048 MB by default (you can lower it to 1024 MB, but heavy pages may crash)
Run timeout	3600 seconds by default

Formats: PNG, JPEG, WebP, PDF. PDF is the page's print layout, not a picture of the screen. WebP can't be taller than 16,383 px, so taller images are made smaller and noted.

Pricing

The Actor is free to use for now. You only pay Apify for the platform usage of your runs (compute and storage). A typical run of a few pages uses a tiny part of the free monthly credit.

Acceptable use

You are responsible for the pages you capture and for any clicks you add. Follow each website's terms and the law.

Not allowed:

• Logins or private pages.
• Getting around age or ID checks, paywalls or CAPTCHAs.
• Harassment, or capturing illegal content.
• Flooding a site with requests.

What the Actor does and does not do:

• It never clicks age checks by itself. If it finds one, the row says "Age check found, not clicked". You may add the button text in "Run these clicks first" at your own responsibility.
• It never types into pages, never logs in, and never solves CAPTCHAs.
• It never opens private, local or internal addresses.

Privacy

• The Actor stores nothing about page content except the images you asked for, in your own run storage.
• Logs show website names only. They never show full links, page text, cookies or HTML.
• It does not use Apify Proxy or any other proxy service, and it does not send your data anywhere else.

FAQ

Why is a page Blocked? The site refused us (for example a 403 error), or the link points to a private address. Some sites block automated tools. Try again later.

Why is a page Cut short? The page is endless (it keeps loading more) or taller than "Max page height". You get the top part. This counts as a delivered screenshot.

Why is a page Partial? The image was saved, but something went wrong. The notes tell you what: a click step was not found, the page was still loading when Max wait ran out, or a cookie banner could not be removed.

Why does a page look different from my browser? Websites show different things by location, cookies, device and time. The Actor opens each page fresh, with no cookies, from a data centre.

How do I add a click step? Open the page in your browser. Find the button or link you want pressed. In "Run these clicks first", add its exact text, for example Show all. Steps run in order before the screenshot.

Why was my page Skipped with "Age check found, not clicked"? The page shows an age check. The Actor never clicks these by itself.

A banner is still in my image. Banner and popup detection is a best guess, and rare designs can be missed. Report the website in the Issues tab and we will look at it.

Support

Found a problem? Open the Issues tab of this Actor. Please include the website name and what you expected.

Coming in version 2

Before/after comparison, dark mode, ad blocking, Standby mode, and a proxy country setting.