All-in-One Facebook Scraper

What does All-in-One Facebook Scraper do?

The most complete Facebook scraper and Facebook data extractor on Apify. Extract data from Facebook Pages, Posts, Events, Groups, Marketplace, Reviews, Comments, Search, Reels, and Ads — all in a single actor. No browser needed, zero proxy costs for public modes, runs on just 512 MB of memory.

Use it as a powerful Facebook API alternative to scrape Facebook posts, extract Facebook page data, scrape Facebook Marketplace listings, scrape Facebook Reels, extract Meta Ad Library data, and build lead lists — all without registering for API access or dealing with rate limits.

🚀 Key Features

• 12 scraping modes in one actor — pages, posts, events, groups, search, reviews, comments, marketplace, reels, ads, all-in-one, data extractor
• HTTP-only architecture — no Playwright, no browser overhead, runs on just 512 MB
• Zero proxy costs — pages, posts, and events work without proxies
• Pay per result — only pay for data you actually receive
• Up to 500 results per input — configurable via resultsPerPage
• GraphQL pagination — authenticated modes tap into Facebook's internal API for deep extraction
• Cookie health monitoring built in — pre-flight check aborts in seconds if cookies have expired (no wasted requests, no escalation of Facebook's abuse score). Optional facebook-monitor mode reports cookie validity, age since last successful preflight, and a cookieless probe on a schedule
• Per-run cost + margin estimate logged at the end of every run — see your effective margin per result without external dashboards
• Structured JSON output — clean, typed fields ready for spreadsheets, databases, or integrations
• 40% cheaper than competing Facebook scrapers on Apify Store

💰 Pricing

Pay-per-result pricing based on scrape mode. No platform fees beyond Apify compute.

Mode	What you get	Free	Starter	Scale	Business
Pages Scraper	Page info, likes, category, verified status	$7.20/1K	$5.40/1K	$4.32/1K	$3.24/1K
Posts Scraper	Post text + engagement metrics	$3.00/1K	$2.25/1K	$1.80/1K	$1.35/1K
Events Scraper	Event name, date, location, RSVP counts	$7.80/1K	$5.85/1K	$4.68/1K	$3.51/1K
Groups Scraper	Group posts with engagement metrics	$3.00/1K	$2.25/1K	$1.80/1K	$1.35/1K
Search Scraper	Find pages by keyword	$7.20/1K	$5.40/1K	$4.32/1K	$3.24/1K
Reviews Scraper	Page reviews and ratings	$1.50/1K	$1.13/1K	$0.90/1K	$0.68/1K
Comments Scraper	Post comments with author info	$1.50/1K	$1.13/1K	$0.90/1K	$0.68/1K
Marketplace Scraper	Listings, prices, sellers, location	$3.00/1K	$2.25/1K	$1.80/1K	$1.35/1K
Reels Scraper	Reel metadata, captions, duration, owner	$3.00/1K	$2.25/1K	$1.80/1K	$1.35/1K
Ads Scraper	Meta Ad Library creatives, dates, platforms	$5.00/1K	$3.75/1K	$3.00/1K	$2.25/1K
All-in-One	Combined modes in one run	$3.00/1K	$2.25/1K	$1.80/1K	$1.35/1K
Data Extractor	Lightweight all-in-one, cheaper billing	$2.00/1K	$1.50/1K	$1.20/1K	$0.90/1K

Up to 40% cheaper than competing Facebook scrapers on Apify Store at every pricing tier.

📊 What Data Can You Extract?

📄 Facebook Page Scraper (no login needed)

• 🏷️ Page name, title, and Facebook ID
• 📂 Category (e.g., "Government organization", "Restaurant", "Software")
• 👍 Likes count, followers count (parsed from "X followers" visible text), talking-about count, check-ins
• ✅ Verified status (via the show_verified_badge_on_profile signal — matches the blue check you see on the page)
• 📍 Location, address, city — populated for pages with a single physical HQ; null for multi-location brands that don't publish an address (Starbucks, McDonald's, etc.)
• 📝 About description (with HTML entities decoded — emoji render correctly)
• 🖼️ Profile picture URL and cover photo URL
• 🔗 Ad Library ID and confirmed page owner (via FB's Ad Library disclosure)
• 🌐 Direct page URL

📝 Facebook Posts Scraper (login cookies required for full fields)

• 💬 Full post text content
• ❤️ Reaction counts per type (Like, Love, Haha, Wow, Sad, Angry) + total
• 💬 Comment count, share count
• 🕐 Publish timestamp
• 📸 Media array (photos/videos with URLs)
• 🆔 Feedback ID, Story ID, Post ID
• 💭 Top comments with author name, text, created time, and likes count — pulled from the post permalink via the unified_reactors key
• 📄 Source page name and URL

Permalink-enrichment scope — the per-permalink fields (topComments, viewsCount on video posts, commentCount, media-fallback) are populated for the first 5 posts per page (the newest, highest-value tier). Posts beyond that come back with feed-level fields only (text, reactions per type, shares, time, postId, url). This bound is intentional — uncapped enrichment burns 1-2 MB Evomi per post and rarely changes the dataset's value. Increase resultsPerPage to scale total post volume; the first-5 enrichment cap stays.

📅 Facebook Events Scraper (no login needed)

• 🎪 Event ID, name, and full description
• 📆 Start/end timestamps + human-readable dayTimeSentence
• 👤 Event organizer name
• 🌐 Online / in-person flag
• 📍 Venue name and city
• 👥 usersResponded (combined going + interested count — Facebook merged these in 2026)
• 🎟️ Ticket URL — note: Facebook stopped populating event_buy_ticket_url even for ticketed events, so this field is now almost always null
• ⚠️ goingCount / interestedCount — partially populated. Facebook removed these keys from most listing-page responses in 2026, but a subset of events still surface them (best-effort). For a reliable combined count use usersResponded (going + interested).

👥 Facebook Group Scraper (requires login cookies)

• 📌 Group title and URL
• 👤 Author name and author ID
• 💬 Full post text
• 🔗 Direct post URL
• ❤️ Reactions, comments, shares
• 🕐 Publish timestamp
• 🆔 Post ID
• 💭 Top comments array
• ℹ️ Works on open/public groups. Closed/members-only groups return 0 posts (expected behavior — the authenticated account must be a member to see posts).

🔎 Facebook Search Scraper (requires login cookies)

• 🏷️ Page name, category, about text
• 👍 Likes and followers count
• 📍 Location
• 🖼️ Profile picture URL
• 🔗 Direct page URL

Memory note: The Search scraper auto-adapts to the available container memory. At the default 512 MB, basic fields (page name, URL, profile picture, verified status) and enriched fields (category, about, location) are returned for the full result set. Bump to 1024 MB if you push resultsPerPage to very large values and see memory pressure in the logs.

Search quality upgrade (2026-04-24): the enrichment step now routes profile.php?id=X micro-pages through the anonymous session pool. Facebook serves those small profile pages as a stripped shell to authenticated sessions, but returns the full OG+meta HTML anonymously — so category, about, followersCount, and location now populate correctly on generic profile.php URLs (e.g. tiny coffee shops, single-person pages) where they previously came back null.

⭐ Facebook Reviews Scraper (requires login cookies)

• 👤 Reviewer name
• 👍 Recommended (yes / no)
• 💬 Full review text (links reviewer ↔ body via a 30K-char context window; handles names with middle initials like "Dock J. Perry" and multi-locale action verbs — recommends, raadt, recomienda, empfiehlt, recommande)
• 📊 Overall rating percentage and total rating count (returned as a separate review-summary item)
• ℹ️ Works best on pages that actually have review bodies. Many pages only collect recommend / don't-recommend clicks without written reviews — in that case the parser returns reviewerName + isRecommended only.
• ℹ️ Pages without reviews enabled (some brand pages return 404 on both /reviews and /ratings paths) are recognized as data-absent and return [] cleanly with no retry. No bandwidth wasted on disabled-review pages.

💬 Facebook Comments Scraper (requires login cookies)

• 👤 Author name
• 💬 Full comment text
• 👍 Likes count — opt-in: set enrichCommentLikes: true in input. When enabled, the scraper fetches each parent post's permalink to backfill likesCount. Default off because each permalink is ~1.2 MB of bandwidth and Facebook only server-renders reaction counts for the post's top-N visible comments — meaning even with enrichment ON, ~50% of feed comments come back null for likesCount. Disable saves ~50% of Comments-mode bandwidth and lifts the run's gross margin substantially. Enable only if your downstream relies on per-comment likes.
• 🕐 Created timestamp
• 🆔 Comment ID + source Post ID (decoded from the comment ID's base64 — 100% reliable)
• 📄 Page name and page URL

🛒 Facebook Marketplace Scraper (works best with cookies)

• 🏷️ Listing title, price, and listing description (full body text) — populated via per-listing detail-page enrichment
• 📦 Item condition — mapped to human-readable strings: "New", "Used - Like New", "Used - Good", "Used - Fair", "Refurbished"
• 💲 Price string + numeric amount + currency code (e.g., USD)
• 📍 Seller name, seller ID, location
• 🔗 Direct listing URL
• 🖼️ Listing images (array with URL + type)
• 🆔 Listing ID and category ID
• ℹ️ Detail-page enrichment runs serially with a humanDelay between fetches (real-shopper cadence) and reliably populates listingDescription, condition, currency, sellerName, and sellerId on ~90%+ of listings per query. Listings beyond the per-query enrichment cap return feed-level fields only (title, price, listingUrl, location, category).

🎬 Facebook Reels Scraper (requires login cookies)

• 🎥 Reel URL and shareable link
• 📝 Caption / text
• ⏱️ Duration (milliseconds)
• 📅 Publish time
• 🖼️ Thumbnail URL
• 👤 Owner name and ID
• 🆔 Reel ID
• ❤️ reactionCount (via unified_reactors, 100% populated when the reel has engagement), commentCount, shareCount
• ℹ️ viewsCount is always null — Facebook does not server-render reel view counts in HTML, even for logged-in accounts (verified 2026-04). There is no workaround without using a headless browser.

📢 Facebook Ads Scraper (requires login cookies)

• 🆔 Ad archive ID and page ID
• 🏢 Page name
• 📝 Ad creative body and title
• 🔗 CTA text and destination link URL
• 🖼️ Ad images and video URLs
• 📅 Start and end dates
• 📊 Active/inactive status
• 📱 Publisher platforms (Facebook, Instagram, Threads, etc.)
• 🌍 Country targeting

🏁 How to Use the Facebook Scraper

Step 1: Create an Apify account

Sign up at apify.com — free tier includes $5 in monthly credits.

Step 2: Open the actor

Go to the All-in-One Facebook Scraper on Apify Store and click Try for free.

Step 3: Configure your input

1. Choose a scrape mode — select from Pages, Posts, Events, Groups, Search, Reviews, Comments, Marketplace, Reels, or Ads
2. Add your targets — enter Facebook page names (e.g., NASA), full URLs (e.g., https://facebook.com/cern), group URLs, or search keywords
3. Set limits — adjust resultsPerPage (default: 50, max: 500)
4. Add login cookies (optional) — required for Groups, Search, Reviews, and Comments modes

Step 4: Run and download results

Click Start and wait for results. Export as JSON, CSV, Excel, or connect to your workflow via API.

🍪 Cookies — Bring Your Own

Important: this actor does NOT include any Facebook cookies for customer runs. When a mode requires authentication, you must paste your own c_user and xs cookies into the loginCookies input field. There is no shared cookie pool, no built-in fallback, and no operator-side cookie injection — every authenticated run uses the cookies you provide, attributing the activity to your Facebook account.

Why this matters:

• Privacy — your scraping activity stays attributed to your own FB account, not someone else's
• Trust — fits Facebook's ToS (the actor cannot run as someone else's identity)
• Quota / rate-limits — your account's anti-bot exposure stays on your account; not pooled with other customers
• Cookie expiry — when your cookies expire, you refresh them; you're not waiting for someone else to update a shared pool

If loginCookies is empty for a mode that requires it (groups, search, reviews, comments, reels, ads), the run errors out immediately with a clear message — by design, to protect both your time and your spend.

How to obtain your cookies

1. Open facebook.com in Chrome and log in
2. Press F12 (Windows) or Cmd+Option+I (Mac) to open Developer Tools
3. Go to Application > Cookies > https://www.facebook.com
4. Copy the values of c_user and xs
5. Paste into the loginCookies field as: c_user=XXXXX; xs=XXXXX

Using multiple cookie sets (recommended for high-volume runs): Enter one account's cookies per line. The scraper will rotate through them automatically if a session gets blocked, extending how long you can scrape before needing fresh cookies:

c_user=REDACTED_FB_COOKIE
c_user=REDACTED_FB_COOKIE
c_user=REDACTED_FB_COOKIE

Cookie expiry: Facebook's xs token rotates every few hours of active use, and the scraper does not persist cookies between runs (the actor treats each run's input as the single source of truth, no background cache). If a run fails with "cookies are invalid" or "login redirect detected," open Facebook in Chrome again, re-extract c_user and xs via DevTools → Application → Cookies, and repaste them into the loginCookies field. The cookie pre-flight check catches this up-front, aborts in seconds, and produces a clear error instead of silently burning requests.

Mode	Login Cookies Required?	Proxy Required?
Pages Scraper	No — cookies suppress likes/follower counts (Facebook serves feed view instead of public page)	Only for brand pages (Apple, Google, SpaceX…)
Posts Scraper	Recommended (enables pagination + full fields)	Only for brand pages
Events Scraper	No	No
Groups Scraper	Yes	Recommended
Search Scraper	Yes	Recommended
Reviews Scraper	Yes	Optional
Comments Scraper	Yes	Optional
Marketplace Scraper	No (but improves results)	Optional
Reels Scraper	Yes	Optional
Ads Scraper	Yes	No

Brand page proxy: Set proxyTier: "residential" (residential proxy included) or proxyTier: "custom" with your own proxy URL + forceProxy: true.

🧭 Human-Like Behavior & Responsible Use

This scraper is engineered to behave like a real browser, not an automation script. Out of the box it:

• Paces requests with randomized, human-like delays between fetches — the delay distribution matches natural reading cadence rather than a fixed interval, so Facebook's anti-abuse models don't see a metronome.
• Applies a session warm-up curve — the first few requests of any new session wait longer than subsequent ones, matching "opened a fresh tab, still reading the first thing before clicking around".
• Occasionally takes longer "think pauses" to break up traffic rhythms the way a real person does when they stop to read a post or check a notification.
• Serializes per-item enrichment (e.g., loading post permalinks or marketplace detail pages one at a time) instead of firing parallel bursts — real browsers physically cannot issue multiple requests on the same millisecond, and doing so is a loud bot fingerprint.
• Keeps browser fingerprints consistent per session — User-Agent, platform hint, and client-hint headers are paired and pinned for the lifetime of a session, the way a real browser sends them. Mixing (e.g. a Windows User-Agent with a macOS platform hint) is an instant giveaway and this scraper never does it.
• Sends a canonical set of navigation headers — Referer, Sec-Fetch-*, cookie consent, locale=en_US, and Upgrade-Insecure-Requests all match a logged-in Chrome tab on macOS/Windows/Linux.
• Rotates sessions frequently so Facebook sees a series of short-lived browser tabs rather than one very long-running one.
• Ships a built-in circuit breaker: at the first sign of throttling (HTTP 429) or a checkpoint challenge, the entire run aborts so the scraper doesn't keep hammering an already-flagged cookie or IP. Retrying after a soft block is one of the biggest contributors to hard suspensions.
• Proactively checks cookie health before any scraping starts. If the cookies are expired or gated, the run aborts in seconds with a clear message — no wasted requests, no escalation of Meta's abuse score against the account.
• Cookies are env-var or input-only — never persisted between runs. Each run reads the cookies from input or actor environment variables and treats them as the single source of truth. If your xs token rotates mid-run the scraper picks it up in-memory for the rest of that run, but never writes it back to any shared cache. This means refreshing your cookies in the input takes effect immediately on the next run without needing to invalidate any cache.
• Enforces optional per-cookie daily + hourly request ceilings (maxDailyRequestsPerCookie, maxHourlyRequestsPerCookie). The run's estimated footprint is checked against the cookie's running counters before any request fires; if the estimate would push the counter past the ceiling, the run aborts immediately. Prevents a misconfigured cron or test runner from burning through an account's headroom.

Safety profiles (safetyProfile input)

Instead of tuning delays and session rotation knobs individually, pick a profile that sets them as a coherent bundle:

Profile	When to use	Character
aggressive	Research accounts you can afford to lose, when speed matters more than safety	Fastest; shorter delays, longer sessions, minimal warm-up
balanced (default)	Everyday scraping on a healthy research/secondary account	Human-like cadence, moderate warm-up, occasional long pauses
safe	Recently-warned accounts, institutional/brand targets, or any workload where stealth beats throughput	Slower delays, shorter sessions, stronger warm-up, more long pauses
first-scrape	Brand-new cookie/IP pair, or immediately after a Facebook suspension is lifted	Ultra-slow onboarding; first requests wait minutes, sessions rotate after very few requests

Tips to scrape responsibly

1. Never scrape with your primary personal Facebook account. Create a dedicated research account and use its cookies. Facebook's anti-abuse models weigh behavioral shifts against a profile's long-term history — a light-history secondary account is dramatically safer than your main one, which has years of organic behavior that a scraping workload will visibly diverge from.

2. Keep batches small at first. Start with 10–25 results per run. Requesting 500 results from one page is much safer than fanning out 25 results × 20 pages in parallel — the first looks like a user reading a timeline, the second looks like a crawler.

3. Space your runs out. Back-to-back runs hitting the same cookie across many different pages in a short window is the single strongest trigger for restrictions. Let runs breathe — if you need continuous data, schedule runs 30–60 minutes apart rather than every minute, and vary the targets between runs.

4. Set per-cookie ceilings on any scheduled or automated workload. A reasonable starting point for a dedicated research account:

   {
       "maxDailyRequestsPerCookie": 500,
       "maxHourlyRequestsPerCookie": 100
   }
   

   Drop both values by ~60% if you must use your main account.

5. Rotate cookies by pasting multiple cookie lines into loginCookies. The scraper swaps between them automatically on a session block, extending how long each account stays healthy.

   c_user=REDACTED_FB_COOKIE
   c_user=REDACTED_FB_COOKIE
   

6. Use the residential proxy tier (proxyTier: "residential") when scraping high-profile brand, institutional, or university pages. Datacenter IPs trip WAF rules more aggressively on those than on smaller pages. The residential proxy is included in the per-result price.

7. Escalate the safetyProfile when something feels off. Go from balanced → safe after any "Circuit breaker tripped" or "Cookie preflight FAILED" message. If you have recently had a Facebook warning or suspension, use first-scrape for the first few runs after the account is cleared.

8. If a run aborts with "Circuit breaker tripped", wait 30–60 minutes before retrying — or swap in fresh cookies. Retrying immediately trains Facebook's anti-abuse model that the cookie is automated and escalates a soft block into a hard suspension.

9. Keep scraping topics aligned with the account's normal browsing when you can. An account whose organic activity is in Dutch about local businesses will look suspicious suddenly scraping dozens of American universities in ten minutes. For scraping work that will diverge sharply from a real person's interests, use a purpose-built account with minimal history.

10. Prefer single-mode runs for sensitive workloads. The all-in-one facebook-scraper mode is convenient but hits many endpoints from the same session in quick succession. For safety-sensitive workloads — or after any prior warnings on the account — run per-mode scrapes one at a time.

Safe starting configurations

Everyday use — a research account in good standing:

{
    "scrapeMode": "facebook-pages-scraper",
    "pages": ["NASA", "cern", "BBCEarth"],
    "resultsPerPage": 10,
    "proxyTier": "residential",
    "safetyProfile": "balanced",
    "maxDailyRequestsPerCookie": 500,
    "maxHourlyRequestsPerCookie": 100
}

Stealth / recently-warned account — slow down and cap hard:

{
    "scrapeMode": "facebook-pages-scraper",
    "pages": ["NASA"],
    "resultsPerPage": 5,
    "proxyTier": "residential",
    "safetyProfile": "safe",
    "maxDailyRequestsPerCookie": 200,
    "maxHourlyRequestsPerCookie": 50
}

First run on a fresh cookie or post-suspension recovery — crawl-speed, tight ceilings:

{
    "scrapeMode": "facebook-pages-scraper",
    "pages": ["NASA"],
    "resultsPerPage": 3,
    "proxyTier": "residential",
    "safetyProfile": "first-scrape",
    "maxDailyRequestsPerCookie": 50,
    "maxHourlyRequestsPerCookie": 20
}

Start small, confirm the results look right, then gradually scale up resultsPerPage and the ceilings. You can also lean on the optional requestDelay input (milliseconds between requests) as an extra throttle on top of the safety profile — set it to 3000 or higher when stealth matters more than speed.

📥 Input Examples

Scrape Facebook Pages

{
    "scrapeMode": "facebook-pages-scraper",
    "pages": ["NASA", "cern", "https://facebook.com/natgeo"]
}

Scrape Facebook Posts

{
    "scrapeMode": "facebook-posts-scraper",
    "pages": ["NASA", "SpaceX"],
    "resultsPerPage": 100
}

Scrape Facebook Events

{
    "scrapeMode": "facebook-events-scraper",
    "pages": ["NASA"],
    "eventSearchQueries": ["tech conference 2026", "music festival new york"]
}

Scrape Facebook Groups

{
    "scrapeMode": "facebook-groups-scraper",
    "groupUrls": ["https://facebook.com/groups/selftaughtprogrammers"],
    "loginCookies": "c_user=XXXXX; xs=XXXXX",
    "resultsPerPage": 50
}

Search Facebook Pages by Keyword

{
    "scrapeMode": "facebook-search-scraper",
    "searchQueries": ["coffee shop new york", "web design agency london"],
    "loginCookies": "c_user=XXXXX; xs=XXXXX"
}

Runs on 512 MB by default — enriched fields (category, about, location) are populated for the full result set. Increase to 1024 MB only if you push resultsPerPage to large values and see memory pressure in the logs.

Scrape Facebook Reviews

{
    "scrapeMode": "facebook-reviews-scraper",
    "pages": ["starbucks"],
    "loginCookies": "c_user=XXXXX; xs=XXXXX",
    "resultsPerPage": 100
}

Scrape Facebook Comments

{
    "scrapeMode": "facebook-comments-scraper",
    "pages": ["NASA"],
    "loginCookies": "c_user=XXXXX; xs=XXXXX",
    "resultsPerPage": 50
}

Scrape Facebook Marketplace

{
    "scrapeMode": "facebook-marketplace-scraper",
    "marketplaceQueries": ["guitar", "laptop", "apartment"],
    "marketplaceLocation": "sanfrancisco",
    "resultsPerPage": 100
}

All-in-One Mode (everything at once)

{
    "scrapeMode": "facebook-scraper",
    "pages": ["NASA"],
    "eventSearchQueries": ["space conference"],
    "groupUrls": ["https://facebook.com/groups/spacenews"],
    "searchQueries": ["space agency"],
    "marketplaceQueries": ["telescope"],
    "loginCookies": "c_user=XXXXX; xs=XXXXX",
    "resultsPerPage": 50
}

📤 Output Examples

Facebook Page Result

{
    "resultType": "page",
    "facebookId": "1504255381069833",
    "pageName": "NASA - National Aeronautics and Space Administration",
    "pageUrl": "https://www.facebook.com/NASA",
    "pageTitle": "NASA - National Aeronautics and Space Administration",
    "category": "Government organization",
    "likesCount": 28644699,
    "followersCount": 28000000,
    "talkingAboutCount": 545254,
    "checkinsCount": null,
    "isVerified": true,
    "isBusinessPageActive": null,
    "about": "Explore the universe and discover our home planet. There's space for everybody. ✨",
    "location": null,
    "address": null,
    "city": null,
    "confirmedOwner": "NATIONAL AERONAUTICS AND SPACE ADMINISTRATION",
    "adLibraryId": "1504255381069833",
    "profilePicUrl": "https://scontent...",
    "coverPhotoUrl": "https://scontent...",
    "scrapedAt": "2026-04-20T08:36:41.222Z"
}

Notes: NASA has no single physical HQ shown on its page → address/city/location are null (this is correct — multi-location brands never show an address). Facebook strips isBusinessPageActive to false for all pages from public HTML, so the scraper returns null rather than a misleading false. For a single-HQ page (e.g. L.L.Bean) these location fields do populate.

Facebook Post Result

{
    "resultType": "post",
    "pageName": "NASA",
    "pageUrl": "https://www.facebook.com/NASA/",
    "postText": "Artemis II crew members are preparing for humanity's return to the Moon...",
    "reactionCount": 45200,
    "commentCount": 1830,
    "shareCount": 8400,
    "publishTime": "2026-03-25T14:30:00.000Z",
    "feedbackId": "10158...",
    "storyId": "S:_I54971236771:...",
    "scrapedAt": "2026-03-27T12:00:00.000Z"
}

Facebook Event Result

{
    "resultType": "event",
    "eventId": "1166242531190119",
    "eventName": "NASA Science Live: Climate Edition - Rising Heat",
    "eventUrl": "https://www.facebook.com/events/1166242531190119/",
    "startTimestamp": "2024-05-06T22:30:00.000Z",
    "endTimestamp": null,
    "dayTimeSentence": "Fri, Jun 28, 2024",
    "eventCreator": "NASA - National Aeronautics and Space Administration",
    "isOnline": true,
    "eventDescription": "From soaring temperatures affecting agriculture and the ocean, to the escalating frequency and intensity of heat waves...",
    "placeName": "Online event",
    "placeCity": null,
    "usersResponded": 105,
    "goingCount": 1240,
    "interestedCount": 780,
    "ticketUrl": null,
    "scrapedAt": "2026-04-20T08:36:44.709Z"
}

Notes: Facebook partially deprecated goingCount / interestedCount in 2026. A subset of events still surface them (as in this example), but most listing-page responses only expose the combined usersResponded count — treat the separate counts as best-effort and usersResponded as the reliable key. ticketUrl is null for this event because event_buy_ticket_url is now stripped from FB HTML even for ticketed events.

Facebook Group Post Result

{
    "resultType": "group-post",
    "groupTitle": "Self-Taught Programmers",
    "groupUrl": "https://www.facebook.com/groups/selftaughtprogrammers",
    "authorName": "Jane Smith",
    "authorId": "100012345678",
    "postText": "Just landed my first developer job after 8 months of self-study!",
    "reactionCount": 342,
    "commentCount": 89,
    "shareCount": 12,
    "publishTime": "2026-03-26T10:15:00.000Z",
    "topComments": [
        { "authorName": "Mike Chen", "commentText": "Congrats!", "likesCount": 15 }
    ],
    "scrapedAt": "2026-03-27T12:00:00.000Z"
}

Facebook Review Result

{
    "resultType": "review",
    "pageName": "starbucks",
    "reviewerName": "Paqui Marques",
    "isRecommended": true,
    "reviewText": "Great coffee and friendly staff. Love the atmosphere!",
    "scrapedAt": "2026-03-27T12:00:00.000Z"
}

Facebook Marketplace Listing Result

{
    "resultType": "marketplace-listing",
    "listingId": "26647526498203861",
    "listingTitle": "Mitchell Acoustic Guitar with Padded Gig Bag - $50 (mission district)",
    "listingPrice": "US$ 50",
    "amount": "50.00",
    "currency": "USD",
    "listingUrl": "https://www.facebook.com/marketplace/item/26647526498203861/",
    "sellerName": "Aaron Stein",
    "sellerId": "1333440687",
    "location": "San Francisco",
    "listingDescription": "Excellent condition, barely played. Comes with hard case and extra strings.",
    "condition": "Used - Like New",
    "category": "676772489112490",
    "media": [{ "url": "https://scontent...", "type": "photo" }],
    "scrapedAt": "2026-04-20T17:08:39.451Z"
}

Facebook Reel Result

{
    "resultType": "reel",
    "reelId": "893935976598019",
    "reelUrl": "https://www.facebook.com/reel/893935976598019",
    "caption": "A huge thank you to our Nutella community...",
    "duration": 10000,
    "thumbnailUrl": "https://scontent...",
    "ownerName": "Nutella",
    "ownerId": "100064593557699",
    "publishTime": "2026-02-27T13:11:16.000Z",
    "reactionCount": 6541,
    "commentCount": 108,
    "shareCount": 96,
    "viewsCount": null,
    "pageName": "nutella",
    "pageUrl": "https://www.facebook.com/nutella",
    "feedbackId": "ZmVlZGJhY2s6...",
    "scrapedAt": "2026-04-20T17:15:00.000Z"
}

Notes: viewsCount is always null on Reels — Facebook does not server-render view counts in reel HTML for any account (verified via Chrome DevTools, 2026-04).

Facebook Comment Result

{
    "resultType": "comment",
    "authorName": "Susan Walker",
    "commentText": "@L.L. Bean My 83-year old mother has been wearing Bean since she was a child...",
    "likesCount": 278,
    "createdTime": "2026-04-10T12:08:19.000Z",
    "commentId": "Y29tbWVudDoxMzM1NjMyNzE4NjEyNzc3XzIzNTg0NjgwMTEzMzAyMDg=",
    "postId": "1335632718612777",
    "pageName": "llbean",
    "pageUrl": "https://www.facebook.com/llbean",
    "scrapedAt": "2026-04-20T17:18:00.000Z"
}

Notes: postId is decoded from the comment ID (base64 of comment:{postId}_{commentId}) — 100% reliable. likesCount is null by default — set enrichCommentLikes: true in input to fetch each parent post's permalink and backfill the count. With enrichment ON, likesCount populates for comments that appear in the permalink's top-N visible comments; lower-ranked feed comments still come back null because Facebook only server-renders reaction counts for the top tier. The opt-in saves ~50% of Comments-mode bandwidth when off.

🌐 Proxy Options — Scraping Brand Pages (Apple, Google, SpaceX)

Facebook blocks Apify's datacenter IPs on high-profile brand pages. When you see "Checkpoint triggered" in the logs, your cookies are fine — the IP is the problem. Two options:

Option A — Built-in residential proxy

No proxy account needed. Just flip one switch and the actor routes through a pre-configured residential proxy. Residential proxy is included.

How to configure:

{
    "proxyTier": "residential"
}

That's it. All modes automatically route through residential IPs.

---

Option B — Bring your own residential proxy

Already have a residential proxy subscription? Paste the URL into the actor:

1. Get a residential proxy URL from your provider (format: http://user:pass@proxy.host:port)
2. In actor input, set Proxy Tier → Bring your own
3. Paste the URL into the Custom Proxy URL field
4. Enable Force Proxy on All Modes → true

{
    "proxyTier": "custom",
    "customProxyUrl": "http://username:password@your-proxy-host:port",
    "forceProxy": true
}

Note: Apify's default "Proxy Configuration" editor has been removed from this actor. Apify Residential Proxy is not supported (its pricing makes the actor's unit economics impossible). Use either the built-in residential tier (proxyTier: "residential") or your own URL via customProxyUrl.

---

When do you need proxy?

Page type	Example	Needs proxy?
Normal pages	NASA, CERN, BBC Earth	❌ No
Brand pages	Apple, Google, SpaceX, Amazon, Microsoft	✅ Yes — residential
Authenticated modes	Groups, Search, Reviews, Comments	✅ Recommended

---

❓ Frequently Asked Questions

Do I need login cookies?

Login cookies (c_user + xs) are required for Groups, Search, Reviews, Comments, Reels, and Ads modes. They are recommended for Posts mode (enables GraphQL pagination and full reaction breakdown). Do not use cookies for Pages mode — when cookies are present, Facebook serves the authenticated feed view which omits og:description, likes count, and follower count entirely. The Events scraper works without cookies. The Marketplace scraper works without cookies but returns more results with them.

What's the difference between modes?

Each mode targets a specific type of Facebook data:

Mode	Best for
Pages	Getting page metadata — likes, followers, category, about text
Posts	Extracting post content with engagement metrics
Events	Finding events from pages or by keyword search
Groups	Reading group discussions and member posts
Search	Discovering pages by keyword (lead generation)
Reviews	Collecting customer feedback and ratings
Comments	Analyzing post engagement and audience sentiment
Marketplace	Monitoring product listings and pricing
Reels	Extracting reel metadata, captions, duration, and engagement
Ads	Browsing Meta Ad Library creatives and ad details
All-in-One	Running multiple modes in a single run
Data Extractor	Same as All-in-One with cheaper per-result billing

How many results can I get?

Up to 500 results per input (per page, group, or search query). Set the resultsPerPage parameter to control this. There is no hard cap on total results — simply add more inputs to scale. For example, 10 pages x 500 posts each = 5,000 posts in one run.

Can I use the API?

Yes. Every Apify actor has a full REST API. You can trigger runs, pass input, and fetch results programmatically:

curl "https://api.apify.com/v2/acts/get-leads~all-in-one-facebook-scraper/runs" \
  -X POST \
  -H "Content-Type: application/json" \
  -H "Authorization: Bearer YOUR_API_TOKEN" \
  -d '{
    "scrapeMode": "facebook-posts-scraper",
    "pages": ["NASA"],
    "resultsPerPage": 50
  }'

See the Apify API documentation for Python, Node.js, and other client libraries.

Is it legal to scrape Facebook?

This Facebook scraper accesses only publicly available data that anyone can view without logging in (for unauthenticated modes). Authenticated modes access data visible to logged-in users. Always ensure your use case complies with applicable laws and regulations. Apify provides the infrastructure; you are responsible for how you use the extracted data. Read more about web scraping legality.

Can I integrate with Make, Zapier, or Google Sheets?

Yes. Apify has native integrations with popular automation tools:

• Make (Integromat) — trigger scraping runs and process Facebook data in Make scenarios
• Zapier — connect Facebook data to 5,000+ apps
• Google Sheets — export results directly to a spreadsheet
• Slack — get notifications when scraping completes
• Airbyte — sync data to your warehouse
• Webhooks — send results to any HTTP endpoint in real time

Does it work with MCP (Model Context Protocol)?

Yes. This actor is fully compatible with the Apify MCP Server. Connect it to AI agents and LLM workflows so tools like Claude, ChatGPT, and custom AI assistants can trigger Facebook scraping runs, process results, and incorporate Facebook data into their responses.

Why is this scraper HTTP-only? Can it handle dynamic content?

Facebook serves most public data in server-rendered HTML and via internal GraphQL endpoints. This scraper parses both — no headless browser needed. This means:

• Faster execution — no browser startup or page rendering delays
• Lower memory — runs on 512 MB vs. 1-4 GB for browser-based scrapers
• Lower cost — no compute overhead from Playwright or Puppeteer
• Zero proxy costs — most modes don't need proxies at all

Which fields are always null and why?

Facebook doesn't expose every field in HTML. The scraper follows a strict null-honesty policy: a field is null only when FB genuinely doesn't carry the data. Ground-truth verified by inspecting FB HTML in a logged-in Chrome session.

Field	Mode(s)	Reason
viewsCount	Reels	Reel page HTML has zero occurrences of video_view_count / play_count / view_count — FB doesn't show view counts for reels
goingCount, interestedCount	Events	Partially populated. Facebook removed the separate keys from most listing-page responses in 2026, but a subset of events still surface them. Combined usersResponded (going + interested) is the reliable key.
ticketUrl	Events	Even ticketed concerts have "event_buy_ticket_url":null — FB stopped populating this key
address, city, location	Pages, Search	Multi-location brands (Starbucks, McDonald's, Walmart…) don't have a single address on FB. Single-HQ pages (L.L.Bean, Zappos, Harley-Davidson) do populate these fields
isBusinessPageActive	Pages	FB strips this to false for every page in public HTML. Returned as null instead of a misleading false
checkinsCount	Pages	Only pages with physical check-in locations have this count
likesCount on Comments	Comments	null by default — opt in via enrichCommentLikes: true to backfill the count for top-N visible comments. Lower-ranked feed comments still come back null even with enrichment on (FB only server-renders reactions for the permalink's top tier).
likesCount on topComments[]	Posts	FB only server-renders reaction counts for the permalink's top-N visible comments
comments, topComments	Posts	Empty when a post has zero user comments (content-dependent, not a parser issue)
viewsCount on non-video posts	Posts	Only video posts expose a view count

What if a mode isn't returning data?

Check the run logs for specific error messages. Common issues include:

• Login cookies expired — re-extract c_user and xs from your browser and paste them into the loginCookies field
• Checkpoint triggered (high-traffic brand pages) — pages like Apple, Google, and SpaceX require a residential proxy. Either set proxyTier: "residential" (built-in) or set proxyTier: "custom" + paste your own URL into customProxyUrl, then enable forceProxy: true. The logs will show "Checkpoint triggered" to distinguish this from cookie expiry.
• Page doesn't exist — verify the page name or URL is correct
• Rate limiting — reduce resultsPerPage or add delays between runs via requestDelay
• Missing input — ensure you've filled in the correct input field for your chosen mode
• Non-English like counts — if your Facebook account is set to a non-English language, the scraper automatically enforces English via a locale=en_US cookie. If counts are still null, try switching your Facebook account language to English

🩺 Health Monitoring & Per-Run Telemetry

facebook-monitor mode — scheduled health probe

Run with scrapeMode: "facebook-monitor" to get a one-shot health report on the actor's authentication and proxy paths. Designed for an Apify Schedule (e.g. every 4 hours) so cookie expiry and proxy issues are caught between scraping runs, not during them.

{
  "scrapeMode": "facebook-monitor",
  "alertEmail": "you@example.com",
  "apifyToken": "your_personal_apify_token"
}

The monitor pushes a single health-report item with these fields:

Field	What it tells you
cookieStatus	valid / expired / missing_or_malformed / test_failed — based on a /me redirect probe
cookielessSmoke	ok / no_og_title / http_<status> / error — independently verifies the proxy + Impit path by fetching facebook.com/NASA anonymously and confirming og:title is present. Catches proxy-format breakage and DNS/network issues that have nothing to do with cookie health
cookieAgeHours	Hours since the last successful preflight for this c_user. Stored as a timestamp only in a tiny KV state store (the cookie value itself is never persisted). Lets you see "cookies are 2h old → safe" vs. "16h old → consider rotating" at a glance
totalRuns, successes, failures, failureDetails, totalCost	Last 20 runs of this actor (excluding monitor runs themselves)

If failures > 0 OR cookies are expired, an alert email is sent to alertEmail summarizing the issues.

Per-run cost + margin estimate

Every non-monitor run logs a single structured line at the end:

INFO  Cost + margin estimate {
  "mode": "facebook-comments-scraper",
  "eventName": "comment-result",
  "eventRegistered": true,
  "delivered": 50,
  "elapsedSec": 62.4,
  "apifyCostUSD_est": 0.0027,
  "revenueUSD_gross": 0.075,
  "ownerRevenueUSD_80pct": 0.060,
  "grossMarginApifyOnly": 96.4,
  "netMarginApifyOnly": 95.5,
  "evomi_usd_per_gb": <your residential proxy $/GB rate>,
  "note": "Residential netRx not visible in-process. True gross margin = (revenue − apifyCost − proxy(netRx_GB × $/GB)) / revenue."
}

grossMarginApifyOnly / netMarginApifyOnly exclude the residential-proxy bandwidth cost (which only the Apify run object knows). Combine with netRxBytes from the run object to get the proxy-adjusted true margin: (revenueUSD_gross − apifyCostUSD_est − netRxBytes × <your $/GB rate> / 1024^3) / revenueUSD_gross. A WARN fires automatically if the Apify-only gross margin drops below 50%.

If the run uses a billing event that isn't registered on the actor's Apify Monetization profile (e.g. you copied this actor and haven't yet registered the per-mode events in the Apify Console), the telemetry prints eventRegistered: false and a Revenue leak: … WARN — the run still produces results, it just won't be charged.

🆚 How It Compares to Other Facebook Scrapers

Feature	All-in-One Facebook Scraper	Typical competitors
Scraping modes	12 modes in one actor	1-3 modes, separate actors
Architecture	HTTP-only (no browser)	Playwright / Puppeteer
Memory requirement	512 MB	1-4 GB
Proxy costs	Zero for most modes	Required
Price per 1K posts	$3.00	$5.00-$8.00
Price per 1K pages	$7.20	$10.00-$15.00
Marketplace support	Yes	Rare
Reviews + Comments	Yes	Rare
MCP server compatible	Yes	Varies
Integrations	Make, Zapier, Sheets, Slack	Varies

Save up to 40% compared to other Facebook scrapers while getting more modes, lower memory usage, and faster results.

🔗 Related Scrapers by get-leads

Building a multi-platform data pipeline? Check out our other scrapers on Apify Store:

• LinkedIn Scraper — profiles, companies, and job postings
• Amazon Scraper — product listings, prices, reviews, and seller info
• Google Maps Scraper — business listings, reviews, and contact details
• TikTok Scraper — videos, profiles, hashtags, and engagement data
• Instagram Scraper — posts, reels, profiles, and follower data

🛠️ Technical Details

• Runtime: Node.js 20 with Impit (Chrome TLS fingerprinting)
• Default memory: 512 MB (min: 256 MB, max: 4,096 MB)
• Concurrency: Auto-managed based on available memory (scales down under pressure)
• Session management: Sticky proxy sessions (same IP reused per cookie session to prevent Facebook xs invalidation). Multiple cookie sets rotate automatically on blocks. Cookies are read from input or actor environment variables on every run — there is no persistent cookie cache. Refreshing the cookies takes effect immediately on the next run.
• Locale handling: Enforces en_US locale automatically to ensure consistent field extraction regardless of the account's language setting
• Error categories: DELETED, RATE_LIMITED, FORMAT_CHANGE, NETWORK, BLOCKED, LOGIN_REQUIRED. DELETED errors (HTTP 404 on a target URL) are now treated as data-absent and do not retry — saves bandwidth on permanently-missing pages (e.g. brand pages that don't have reviews enabled). Transient errors (timeouts, connection resets, HTTP 5xx) retry within the same session.
• Per-run telemetry: every non-monitor run logs a Cost + margin estimate {…} line at the end with delivered, apifyCostUSD_est, revenueUSD_gross, grossMarginApifyOnly, netMarginApifyOnly. A WARN fires when estimated gross margin drops below 50%, or when the run uses a billing event that isn't yet registered on the actor's Apify Monetization profile.
• Output formats: JSON, CSV, Excel, XML, HTML
• Proxy: Not required for pages/posts/events. Recommended for groups/search. Required (RESIDENTIAL) for high-traffic brand pages (Apple, Google, SpaceX) when using forceProxy: true

---

Built and maintained by get-leads on Apify. Found a bug or have a feature request? Open an issue on the Issues tab.