Facebook Public Page Scraper (no login)

Scrapes public Facebook Pages over plain HTTP — no browser (no Playwright/Chromium) and never logs in. No account, no cookies, no email/password. It sends a realistic browser TLS/HTTP2 fingerprint (got-scraping) so logged-out Facebook answers with the full page, then parses the HTML with cheerio. Because it is HTTP-only, it runs fast and cheap on an Apify DATACENTER proxy.

For each page it returns the full public identity plus the recent public posts Facebook still serves to logged-out visitors.

Input

{
  "pageUrls": ["nasa", "https://www.facebook.com/CocaColaUS", "@natgeo"],
  "maxPosts": 20,
  "proxyConfiguration": { "useApifyProxy": true, "apifyProxyGroups": ["BUYPROXIES94952"] },
  "timeoutSecs": 45
}

• pageUrls — usernames or full page URLs (with or without @/https://).
• maxPosts — cap on posts per page (see the note below on the logged-out limit).
• proxyConfiguration — DATACENTER (BUYPROXIES94952) is the cheap default; switch to RESIDENTIAL only if a page comes back blocked.
• timeoutSecs — per-request timeout (15–120s).

Output (one dataset item per page)

{
  "pageUrl": "https://www.facebook.com/nasa/",
  "loggedIn": false,
  "method": "http",
  "resolvedVia": "www.facebook.com",
  "name": "NASA - National Aeronautics and Space Administration",
  "category": "Government organization",
  "followers": "28,731,417",
  "likes": null,
  "intro": "NASA ... 28,731,417 followers · ...",
  "website": null,
  "email": "public-inquiries@hq.nasa.gov",
  "phone": null,
  "profilePic": "https://scontent.../....png",
  "isVerified": true,
  "postsFound": 1,
  "posts": [
    {
      "text": "Tune in now to watch ...",
      "timestamp_iso": "2026-09-26T16:59:18.000Z",
      "permalink": "https://www.facebook.com/NASA/posts/pfbid02...",
      "reactions": 308,
      "comments": 28,
      "shares": 14,
      "media": ["https://scontent.../image.jpg"]
    }
  ],
  "note": "OK: identity + 1 public post(s) ...",
  "attempts": [ ... ]
}

How it works

• www.facebook.com is the primary host: the logged-out desktop HTML embeds the page identity in its og:/meta tags AND the recent public post(s) as JSON inside <script type="application/json"> Relay caches (real message text, permalink, creation_time, reaction/comment/share counts). That JSON is parsed directly — no DOM rendering needed.
• m.facebook.com is a light identity-only fallback used if www yields nothing.
• mbasic.facebook.com is intentionally not used — it now hard-redirects logged-out visitors to a login page, so it yields no public data.

Important limitation (be honest)

Logged out, Facebook only exposes a small number of recent public posts — in practice the pinned/top post (typically 1–4 posts). Getting the full historical feed of a page requires a logged-in session, which this actor never uses. When Facebook exposes no posts, the actor returns identity only with posts: [] — it never fabricates or pads posts.