Website Screenshot API: Full-Page PNG & Visual Diff

Most screenshot APIs hand you a picture and forget it. Then you build the rest yourself: a storage bucket for old images, a diff tool, a scheduler, and a rule for what counts as "changed". This Actor does that part in the same run: a full-page screenshot, the share of changed pixels against the previous screenshot of the same URL, and a diff image with the changes painted red. The new screenshot becomes the baseline for the next run - automatically.

One call = one page. Nothing to install, nothing to store on your side.

The pain this removes

The usual story	What happens here
Screenshot taken, change detection left to you	changed, changedRatio and diff.png come back in the same run
Baselines on your disk, or silently mixed with other users'	A named store screenshot-baselines inside your own Apify account, keyed by URL + viewport + format
Blank or half-rendered images on JavaScript-heavy sites	Real Chromium, waits for the load event plus waitFor ms for lazy images
A monthly subscription for a handful of checks	$0.05 per page, no subscription, no minimum
"It worked" with nothing to prove it	Every run is a dataset row: size, dimensions, timestamp, keys, change ratio
Mobile and desktop shots overwriting each other	Viewport and format are part of the baseline key, so each keeps its own history

What you get

Output	Where
Change ratio, baseline key, capture metadata	Dataset (Run result)
Screenshot (PNG or JPEG)	Key-value store, screenshot.png
Diff image, red = changed pixels	Key-value store, diff.png - only when something changed

Example dataset item:

{
  "url": "https://example.com",
  "changed": false,
  "changedRatio": 0.0,
  "firstRun": false,
  "sizeBytes": 71152,
  "width": 1440,
  "height": 900,
  "fullPage": true,
  "screenshotKey": "shot_2f1a6c1e8b3f0d92",
  "checkedAt": "2026-11-04T12:15:00+00:00"
}

When the page differs from the baseline, changed is true, changedRatio holds the share of changed pixels (for example 0.0853 = 8.53 %) and an extra record diffKey points to the diff image.

Who uses it

• Price and content monitoring - watch a product page, a tariff table or a hotel page for edits.
• Competitor landing pages - a silent redesign shows up as a change ratio, not as a surprise next month.
• Terms of service and legal pages - a timestamped image of what the text said on a given day.
• Status pages and dashboards - a picture is the proof when the status API is down.
• Deploy checks - your own site after a release, with an alert when the layout breaks.
• QA of marketing pages across viewports, without a browser in CI.

Input

Field	Default	Meaning
url	-	Page address. example.com is accepted and upgraded to https://.
fullPage	true	Whole scrollable page, or only the visible screen.
width / height	1440 / 900	Viewport in CSS pixels.
format	png	png for pixel-perfect detail, jpeg for smaller files.
diffPixels	12	How much a colour channel must change (0-255) for a pixel to count. Higher ignores soft noise, lower catches subtle shifts.
waitFor	1000	Extra milliseconds after load, for lazy images and animations.
resetBaseline	false	Forget the stored screenshot for this URL and treat this run as a new baseline.

Use it from code

curl -X POST "https://api.apify.com/v2/acts/YOUR_USER~website-screenshot-visual-diff/run-sync-get-dataset-items?token=YOUR_TOKEN" \
  -H "Content-Type: application/json" \
  -d '{"url":"https://example.com","fullPage":true}'

The response is the dataset item above; the images are in the run's key-value store under the returned keys. Works with Python, Node, Zapier, Make and n8n through the same endpoint.

Use it from an AI assistant (MCP)

The Actor is callable as an MCP tool, so Claude, Cursor or any MCP client can check a page directly: "compare this page with the version from last week" - the assistant runs the Actor and reads the change ratio instead of guessing.

Where the baselines live

In a named key-value store inside your own Apify account (screenshot-baselines). They survive between runs, they are billed to you at Apify storage prices, and nobody else's baselines are mixed with yours.

Pricing

$0.05 per screenshot page - one event covers the screenshot and its comparison with the previous run, no matter how large the page is. No subscription, no minimum. On Apify's free plan the run is covered by the included monthly usage credit, so no card is needed until that credit runs out.

Tips for stable results

• Pages with carousels, live counters or ads change every minute - raise diffPixels to 20-30 to ignore them.
• Cookie banners appear on the first visit only; run twice with resetBaseline: false to get a clean baseline.
• For a long page with lazily loaded images, raise waitFor to 3000 ms.
• To watch many pages, run the Actor once per URL via API, a schedule, or an Apify task.

Limitations

• Pages behind a login or a paywall cannot be captured (no authentication is sent).
• Very long or heavy pages may hit the run timeout; increase the timeout and memory in the run settings.
• The diff is pixel-based, so a 1-pixel layout shift counts as a change across the whole page.

Feedback

A page that renders badly, or a site that refuses to load? Open an issue on the Actor page with the URL - that is the fastest way to get it fixed.

Spending cap

If you do not set a spending cap, Apify applies its own default (about $4.63 per run). When that cap is reached the run stops early: the remaining items are skipped and you are charged only for what was actually delivered. For a full pass, set maxTotalChargeUsd in the run options.