Facebook Page Photo Gallery & Engagement Scraper

Collect public Facebook photos from pages, profiles, albums, tagged collections, and single-photo links. Each row carries the full-size image and thumbnail, caption, upload time, reaction breakdown, comment and share counts, tagged people, owner/page context, album, and post link. No login, cookies, browser, or proxy setup.

What this actor does

• Walks a page or profile's Photos tab or its Tagged tab; add both URLs to collect both
• Expands albums, a page's Albums list, or a direct album URL into its photos
• Accepts single-photo permalinks and photo paths
• Follows Facebook's cursor pagination 8 photos at a time, de-duplicates, and keeps going past the first 8-photo batch
• Applies newerThan / olderThan on Facebook's side for page and tagged tabs (Unix-second query arguments)
• Opens each photo for caption, timestamp, reactions (7 types), comments, shares, tagged people, album, post, and owner
• Copies images to durable storage by default and keeps the signed original plus its expiry

Input

Field	Type	Default	Description
startUrls	array	–	Page, profile, /photos, /photos_by, /photos_stream, /photos_of, /photos_albums, album, photo permalink, or bare handle
resultsLimit	integer	100	Global cap across all Start URLs
newerThan	string	–	Keep photos posted on or after this YYYY-MM-DD
olderThan	string	–	Keep photos posted on or before this YYYY-MM-DD
photoTab	select	uploaded	For page/profile URLs: uploaded or tagged
includeAlbums	boolean	false	Also walk every album on a page/profile
includeDetails	boolean	true	Open each photo; turn off for a fast grid-only export. Album date bounds still read only the timestamp needed for filtering
storeImages	boolean	true	Store permanent image and thumbnail copies in Apify storage

proxyConfiguration is no longer an input. The actor selects its own connection path; old saved tasks that still send the key are accepted and the key is ignored.

Examples

A page's recent photos:

{
  "startUrls": [{"url": "https://www.facebook.com/humansofnewyork/photos"}],
  "resultsLimit": 50
}

Tagged photos of a page:

{
  "startUrls": [{"url": "https://www.facebook.com/nike"}],
  "photoTab": "tagged",
  "resultsLimit": 100
}

Every album on a page:

{
  "startUrls": [{"url": "https://www.facebook.com/nasa/photos_albums"}],
  "resultsLimit": 300
}

A date window:

{
  "startUrls": [{"url": "https://www.facebook.com/nationalgeographic/photos"}],
  "newerThan": "2025-01-01",
  "olderThan": "2025-12-31",
  "resultsLimit": 200
}

A single photo:

{
  "startUrls": [{"url": "https://www.facebook.com/photo/?fbid=1547667900257527"}]
}

Output per photo

Legacy fields keep their names: photoId, photoUrl, imageUrl, caption, altText, timestamp, likesCount, commentsCount, reactionsByType, reactionsTotal, shareCount, taggedPeople, authorName, authorUrl, inputUrl, scrapedAt.

Added fields include imageUrlOriginal, imageExpiresAt, imageWidth, imageHeight, thumbnailUrl, thumbnailUrlOriginal, ocrText (Facebook's accessibility caption, not a second OCR pass), createdTime, likeCount…angryCount, uploaderId/Name/Url, pageId/Name/Url/Followers, albumId, albumName, postId, postUrl, sourceType, and detailsStatus.

likesCount keeps its old meaning (total reactions); likeCount is the LIKE reaction only. Empty values are omitted. A photo without a caption simply has no caption key; a zero count is kept when Facebook reports zero. With includeDetails off, album rows have only the grid thumbnail and thumbnail dimensions; imageUrl requires details. Album date bounds still work in this mode: the actor reads just the timestamp and keeps the row grid-only.

Pagination and limits

Facebook returns 8 photos per collection request. The actor follows the cursor until the source ends, resultsLimit is reached, three consecutive requests add no new ids, or a duplicate-heavy safety budget is hit. A collection also stops after 2,000 pages (about 16,000 photos) as a safety cap. When a target stops early, the status message notes the shortfall; the log may not name the exact stop cause. No 1,000-photo or 5,000-photo artificial cap is applied.

Image links and storage

Facebook's CDN links are signed and expire. With storeImages on (default), each image and thumbnail is copied to Apify storage; when a copy succeeds, imageUrl / thumbnailUrl become permanent public links. imageUrlOriginal and imageExpiresAt preserve the source. If a copy fails, the signed source link is kept and the run continues. With storeImages off, the signed links are returned as-is and the expiry is still recorded.

What this actor cannot collect

• Photo location — Facebook does not expose a place/location for photos on these surfaces
• Multiple image resolutions — the photo viewer exposes one image URI plus width/height
• Private profiles, login-only pages, and groups — only what a logged-out visitor sees
• Comment bodies or reactor identities — counts only; use the comments actor for threads
• Videos and Reels — the Photos collections contain photos; use the posts/reels actors for video
• One row per album — album metadata is attached to its photos; an albums-list URL expands into photos

FAQ

Do I need a Facebook account? No. The actor reads only public, logged-out content and never asks for cookies or credentials.

Why did a target return no photos? It may not exist, be private, or hide its Photos tab from logged-out visitors. The final status message names failed targets.

Why are some captions or owner names missing? Facebook genuinely has no caption for some photos, and occasionally omits the uploader's display name. Empty fields are omitted rather than filled with guesses.

What does detailsStatus mean? Complete means the photo page was read; GridOnly means details were turned off; Unavailable means the photo page could not be read, so grid fields were kept.

Why do some images need a slower request? Gallery pages are free to read. Individual photo pages are often served as a stripped page to datacenter IPs, so the actor automatically retries the detail request through a residential connection. The gallery does not use that path.

How many photos can I get? As many as the page exposes to a logged-out visitor, up to resultsLimit. A collection stops after 2,000 pages (about 16,000 photos) even if more remain. Thousands of photos have been walked from one page; duplicate-heavy pages may stop at the safety budget and the status message notes the shortfall.

Can I filter by date? Yes. Use YYYY-MM-DD. Page and tagged tabs send the window to Facebook. Albums have no server-side date filter, so each photo has to be opened and checked. This works even with includeDetails off: the actor reads only the timestamp needed for the bound, keeps the row grid-only, and stops with a clear status if the window is not reached within the scan budget. An album date run costs one detail request per photo examined.

Are image links permanent? With storeImages on, yes, when the copy succeeds; a failed copy leaves the signed source link. With it off, Facebook's signed links expire.

Is this affiliated with Facebook? No. This is an independent third-party tool that reads public pages; it is not affiliated with or endorsed by Meta.

Data source

Public facebook.com pages, profiles, photo albums, and photo permalinks, read without login. The actor does not bypass private content or login walls.