Facebook Posts Scraper

Collect posts from public Facebook Pages and anonymously accessible profiles. Get text, publication dates, authors, engagement counts and available media references in one dataset. No Facebook account or cookies are required.

$0.99 per 1,000 saved posts on Free / Starter, $0.89 on Scale and $0.79 on Business. Platform usage is included: no separate compute, proxy, Actor start or date-filter fee. Twenty saved posts cost $0.0198 at the Free / Starter rate. Check the Pricing tab for the active rate before starting; Free is the account plan, not a free trial.

Use this Actor for content research, publishing activity and engagement comparisons. Choose an independent post limit for each source, or collect a publication window. Requested sources, observed authors and shared originals stay distinct in the results.

First run: one Page, up to 20 posts

1. Open the Input form and paste a public Page or profile link into Page or profile URLs. The prefilled BBC News Page is a starting example.
2. Set Maximum posts per source to 20, then click Start.
3. When processing finishes, open Output to inspect the dataset and download results. Also check the run summary for each source's outcome.

The equivalent JSON input is:

{
  "sourceUrls": ["https://www.facebook.com/bbcnews/"],
  "maxPostsPerSource": 20
}

The default table shows the first observed author and first media preview when available. Full details keeps all observed authors, media, reactions, shared-original context and field issues. A missing preview does not mean the post has no media.

What results look like

This is a synthetic illustration, not a Facebook capture or an Apify Console screenshot. Each row is one post; nested authors or media do not create extra post rows.

Post	Published (UTC)	First author	Reactions	Comments	Shares
An example update with a link.	2026-10-01 12:00	Example Page	125	18	7

Selected fields from the same synthetic row:

{
  "postId": "9000000000000003",
  "createdAt": "2026-10-01T12:00:00Z",
  "text": "An example update with a link.",
  "textLinks": ["https://example.com/article"],
  "authors": [
    {"id": "9000000000000001", "name": "Example Page", "url": "https://www.facebook.com/example/", "avatarUrl": null}
  ],
  "reactionsCount": 125,
  "commentsCount": 18,
  "sharesCount": 7,
  "attachments": [],
  "sharedPost": null,
  "issues": [],
  "sourceUrl": "https://www.facebook.com/example/"
}

textLinks contains observed message destinations. attachments holds available media or link attachments; it is empty in this illustration. sharedPost, when available, contains the original post's own text, authors and media separately from the collected wrapper. The JSON above shows selected fields; complete rows also carry source and input identity.

Collect from several Pages

For a comparison, request up to 100 posts from each source. This example saves at most 200 rows, costing at most $0.198 at the $0.99-per-1,000 event rate if all 200 are saved.

{
  "sourceUrls": [
    "https://www.facebook.com/bbcnews/",
    "https://www.facebook.com/cern/"
  ],
  "maxPostsPerSource": 100
}

Every input is processed independently, including repeated inputs. A post appearing in several sources is saved and billed separately for each input. Fresh runs can repeat posts too.

Research a publication period

{
  "sourceUrls": ["https://www.facebook.com/bbcnews/"],
  "maxPostsPerSource": 200,
  "since": "2026-10-01",
  "until": "2026-10-07"
}

Calendar dates include the whole UTC day. Exact boundaries accept timezone-aware, whole-second timestamps such as 2026-10-01T12:00:00Z or 2026-10-01T15:00:00+03:00. Both bounds are inclusive. Posts with unknown publication dates are retained with an unverifiable_date_window issue; they are not confirmed matches for the chosen period.

Monitor new publications

For each run, set since to your last completed cutoff and until to the end of the new window. Include some overlap and deduplicate saved results by postId; inclusive bounds can return a boundary post again. Advance the cutoff only after that source's window ends complete. limit_reached, incomplete, a partial run or a timeout does not establish coverage: increase the limit or resolve the interruption before advancing. Feed order is not guaranteed newest first. Keep unknown-date rows separate when assessing window coverage.

Checking engagement on older posts is a different job from finding new publications: use a window that includes those older posts to collect their current available counts. The Actor keeps no automatic monitoring state or resume checkpoint.

Input reference

Option	Meaning
sourceUrls	Public Facebook Page/profile URLs; direct post, group and search URLs are unsupported.
sourceIds	Numeric Page/profile IDs as strings, as an alternative to URLs.
maxPostsPerSource	Strict saved-post limit for each input: default 100, allowed 1–10,000; the form starts at 20.
since, until	Optional inclusive publication bounds, as UTC days or timezone-aware timestamps.

Provide 1–1,000 sources across the two source lists combined. URLs are processed before IDs. A permitted limit does not guarantee the run can finish within its timeout or budget.

Output fields

Fields	Contents
postId, url	Collected post identity and observed permalink.
text, textLinks	Available text and observed message links.
createdTime, createdAt	Publication time as Unix seconds and a UTC timestamp.
authors	All observed authors with available IDs, names, profile and avatar URLs.
reactionsCount, commentsCount, sharesCount, videoViewCount	Available engagement counts; unknown values stay null.
reactions	Observed reaction types and counts, which may be a subset.
attachments, sharedPost	Feed-visible media/link references and separate shared-original context.
inputIndex, inputUrl, inputSourceId, sourceId, sourceUrl	Original input position and resolved source identity.
feedbackId, issues	Observed feedback identity and field-level uncertainty.

Media references are not downloaded files. Viewer links, previews and direct media URLs have separate fields; URLs may expire. Album items and video variants are limited to what the feed exposes. No complete album, complete text/link inventory or full archive is promised. Unknown data stays null with issues where applicable; observed zero and empty lists remain distinct.

Exports, API and schedules

Apify provides dataset downloads for formats such as JSON, CSV and Excel, and API access to start runs from another application. These are Apify platform features. Use Apify schedules for recurring runs. A normal schedule reruns the input you supplied: fixed since/until dates do not advance automatically. Update those bounds yourself or through your workflow for moving windows.

Run outcomes and common questions

Where do I check completeness? Read the unbilled run summary, including every input. complete means the available traversal finished, including confirmed zero results; limit_reached means the requested cap was reached. incomplete stopped early; failed / invalid_source could not be collected; skipped was not processed. A successful platform run can contain unsuccessful inputs. An inaccessible source does not establish that it is private, deleted or empty. Other inputs' saved posts remain available.

What survives an interrupted run? Acknowledged rows remain in the dataset. Some inputs may be unfinished, and interruption can leave summary counts behind the retained rows. Check the dataset itself. A new run starts over and may save and bill the same posts again.

Are empty inputs or the summary charged? No. Billing counts saved post rows, including duplicates across independently processed inputs or repeated runs. There is no separate Actor start, publication-date-filter, compute or proxy fee.

Can I collect comment text? This Actor returns available comment counts. Use the separate DataWan Facebook Comments Scraper for comment bodies and optional replies; its runs and charges are separate.

Need help? Open an issue on this Actor with the run link, intended input and expected outcome. Do not include account credentials.