Facebook Comments Scraper

Facebook Comments Scraper — Extract Comments, Replies and Profiles

Facebook Comments Scraper Pay Per Events extracts public comments, nested replies, and commenter profile data from any Facebook post, photo, or reel permalink, and returns them as one flat JSON row per top-level comment — no login, cookie, or Graph API access token required. Unlike scraping frameworks that hand back raw HTML, this Actor returns typed JSON, ready for a spreadsheet, a database, or a model's context window without any parsing. Pricing is pay-per-result: you're billed per comment row actually written to your dataset, never for the run itself. This guide covers every input and output field, plus the deployment patterns teams use to run it in production.

🧭 What Does Facebook Comments Scraper Pay Per Events Do?

Facebook Comments Scraper Pay Per Events reads the public comment thread on a Facebook post, photo, or reel permalink and returns every top-level comment — with its nested replies attached — as a flat, structured dataset row: author name and ID, comment text, timestamp, reaction count, and reply count. No Facebook account, login cookie, or Graph API access token is required — the Actor loads the public permalink logged-out through Apify's residential proxy and paginates Facebook's own GraphQL comment API directly.

• Extracts every top-level comment on a post, photo, or reel, up to commentSettings.maxComments
• Expands nested reply threads per comment, up to commentSettings.maxRepliesPerComment
• Orders output rows deterministically by commentsSortOrder (most_relevant, newest, or all)
• Filters to a date window with onlyCommentsNewerThan (absolute date or relative span like 2 weeks)
• Returns commenter identity fields — profileName, profileId, profileUrl, profilePicture — with every row
• Attaches Facebook Ad Library metadata (pageAdLibrary) when a comment is tied to a Page running ads
• Bulk-processes multiple post URLs from one startUrls list in a single run

⚡ Features & Capabilities

Facebook Comments Scraper Pay Per Events groups its capabilities into three areas: what it extracts per comment, how it fits into the rest of the Scrapio Facebook toolset, and how it bills for results.

Core features

• Scrapes public comments from Facebook post, photo, and reel permalinks, with bulk URL paste supported in startUrls
• Attaches nested reply threads to their parent comment in the reply array, capped by commentSettings.maxRepliesPerComment
• Deterministic output ordering via commentsSortOrder: most_relevant (engagement-ranked), newest (date-descending), or all (Facebook's returned order)
• Date-window filtering via onlyCommentsNewerThan, applied after comments are fetched
• Every row carries author identity (profileName, profileId, profileUrl, profilePicture), comment content (text, date), and engagement (likesCount, commentsCount)
• Ad Library metadata (pageAdLibrary.is_business_page_active, pageAdLibrary.id) attached when a comment's author is a Page with an Ad Library presence
• Always-on Apify residential proxy routing — no manual proxy setup required, regardless of what is selected in proxyConfiguration
• Pay-per-result pricing: one row_result charge per top-level comment row delivered (nested replies inside that row are not billed separately); rows that fail to load are pushed without a charge

Facebook Comments Scraper Pay Per Events within the Scrapio data stack

Facebook Comments Scraper Pay Per Events covers comments and replies on individual post, photo, and reel permalinks. For comment threads inside public Facebook Groups instead of Pages, use facebook-groups-comments-replies-scraper. To discover post URLs at scale by keyword before feeding them into this Actor, use facebook-posts-search-comments-scraper or facebook-page-posts-scraper-with-lead-contact-list. For advertiser contact details behind the Pages running ads that show up in pageAdLibrary, see facebook-ads-library-scraper-advertiser-contact-finder.

Why do developers and data teams scrape Facebook?

Different teams pull Facebook comment data for different reasons — the common thread is that the comment thread, not just the post, is the signal they actually want.

🏢 Social listening, brand monitoring and moderation teams

Brand and community teams monitor comment threads on their own posts, and on competitor or influencer posts, to catch complaints, praise, and moderation-worthy content early. Feed a page's recent post URLs into startUrls, set commentsSortOrder to newest so freshly posted comments surface first, and use onlyCommentsNewerThan (e.g. 1 day) to pull only what happened since the last check. Each row's text, profileName, likesCount, and date fields are enough to flag a spike in negative sentiment, a customer complaint needing a reply, or a comment worth reporting. Because replies are nested in the parent comment's reply array, teams can see whether a page admin already responded before escalating.

📊 AI training data and RAG indexing

Comment text is the highest-information field for both use cases — short, first-person, unstructured natural language, exactly the shape RAG pipelines chunk and embed for retrieval, and exactly what makes useful data for fine-tuning sentiment or intent classifiers. For RAG enrichment, index text alongside profileName, date, and the parent facebookUrl so a retrieved snippet can cite who said what, when, and on which post. For training data, likesCount and commentsCount work as a weak engagement label, while threadingDepth distinguishes an original top-level opinion (0) from a reactive reply (1). Every field returns as a typed primitive straight out of the dataset — no HTML stripping or regex cleanup before it reaches an embedding model or a fine-tuning job.

📱 Competitive and market intelligence

Marketing and competitive-intelligence teams point startUrls at a competitor's or industry page's public posts to track how audiences react — not just what a competitor posts, but how people respond to it. Sorting with commentsSortOrder: most_relevant surfaces the comments Facebook itself ranks as highest-engagement, a fast way to see which claims, offers, or product angles actually land. Tracking likesCount and commentsCount on the same post across repeated runs, using onlyCommentsNewerThan to pull only the delta, shows whether a competitor's engagement is genuinely accelerating or just accumulating slowly.

🔬 Research and academic use

Academic and social-science researchers use public comment threads to study online discourse, sentiment, and community response to news, political, or public-health content, without a Facebook Graph API app review or data-partnership process. Because every row is built only from a publicly reachable post permalink, this Actor is scoped to public-data-only research designs — content behind a login wall, a private group, or a friends-only audience is never reachable and never appears in the output.

🎥 Product and SaaS development

Teams building moderation dashboards, sentiment-monitoring tools, or lead-scoring products on top of Facebook engagement data use this Actor as the ingestion layer: schedule it against a client's own pages, normalize text, likesCount, and profileName into your own schema, and layer your product's scoring or alerting logic on top. Because output fields are stable JSON keys rather than scraped HTML, maintaining that ingestion layer doesn't mean rewriting a parser every time Facebook changes its front end.

🍚 Input Parameters

All parameters below are read directly from .actor/input_schema.json (embedded in actor.json), in schema order, with the same names, types, and defaults the Actor actually uses.

Parameter	Required	Type	Description	Example Value
startUrls	Yes	array of strings	One Facebook post, photo, or reel URL per line; bulk paste from a spreadsheet is supported.	["https://www.facebook.com/humansofnewyork/posts/pfbid0BbKbkisExKGSKuhee9a7i86RwRuMKFC8NSkKStB7CsM3uXJuAAfZLrkcJMXxhH4Yl"]
commentSettings	No	object	Grouped panel for comment limits and sort order (see nested fields below).	{ "maxComments": 100, "maxRepliesPerComment": 2, "commentsSortOrder": "all" }
commentSettings.maxComments	No	integer, minimum 1	Hard cap on top-level comments fetched per URL. Default: 100.	100
commentSettings.maxRepliesPerComment	No	integer, minimum 0	Nested replies fetched per threaded comment. Default: 2.	5
commentSettings.commentsSortOrder	No	string, enum: most_relevant, newest, all	Order of the comment rows in the output. Default: all.	"newest"
onlyCommentsNewerThan	No	string	Absolute date (2024-01-15) or relative window (1 day, 2 weeks, 3 months). Only comments and replies posted after this point are kept. Filtering happens after comments are fetched, so set maxComments generously for narrow windows. Leave blank to disable.	"2 weeks"
proxyConfiguration	No	object	Informational only — the Actor always attaches Apify residential proxy at runtime regardless of this setting. Prefill: { "useApifyProxy": false }.	{ "useApifyProxy": false }

Example input:

{
  "startUrls": [
    "https://www.facebook.com/humansofnewyork/posts/pfbid0BbKbkisExKGSKuhee9a7i86RwRuMKFC8NSkKStB7CsM3uXJuAAfZLrkcJMXxhH4Yl"
  ],
  "commentSettings": {
    "maxComments": 100,
    "maxRepliesPerComment": 5,
    "commentsSortOrder": "newest"
  },
  "onlyCommentsNewerThan": "2 weeks"
}

Supported URL types and input formats

startUrls accepts typical Facebook post, photo, and reel permalinks from facebook.com — anything starting with http is accepted for processing. The Actor then loads that URL as a Facebook permalink page and extracts the doc_id/feedback_id pair its GraphQL comment API needs; if that extraction fails (wrong URL type, removed post, or a page Facebook won't render logged-out), the row for that URL carries an error message instead of comments rather than stopping the run.

Real examples:

• Standard post permalink: https://www.facebook.com/humansofnewyork/posts/pfbid0BbKbkisExKGSKuhee9a7i86RwRuMKFC8NSkKStB7CsM3uXJuAAfZLrkcJMXxhH4Yl
• Bulk input — multiple posts in one run, one per line:

  https://www.facebook.com/humansofnewyork/posts/pfbid0BbKbkisExKGSKuhee9a7i86RwRuMKFC8NSkKStB7CsM3uXJuAAfZLrkcJMXxhH4Yl
  https://www.facebook.com/PageName/videos/1234567890123456
  

• Narrow time-window run — pair a recent post with a relative filter: startUrls set to the post, onlyCommentsNewerThan set to "3 days", and commentSettings.maxComments raised (e.g. 500) since the date filter is applied after the fetch.

📦 Output Format

Each top-level comment is pushed to the dataset as one flat row, with any nested replies carried inside that same row's reply array — replies are never written as separate dataset rows. Rows stream into the Output tab as the run works.

Output for Comments

{
  "facebookUrl": "https://www.facebook.com/humansofnewyork/posts/pfbid0BbKbkisExKGSKuhee9a7i86RwRuMKFC8NSkKStB7CsM3uXJuAAfZLrkcJMXxhH4Yl",
  "commentUrl": "https://www.facebook.com/humansofnewyork/posts/pfbid0BbKbkisExKGSKuhee9a7i86RwRuMKFC8NSkKStB7CsM3uXJuAAfZLrkcJMXxhH4Yl?comment_id=1234567890123456",
  "id": "Y29tbWVudDoxMjM0NTY3ODkwMTIzNDU2",
  "feedbackId": "ZmVlZGJhY2s6MTIzNDU2Nzg5MDEyMzQ1Ng==",
  "date": "2026-06-02T14:08:31+00:00",
  "text": "This story hit hard. Thank you for sharing it.",
  "profilePicture": "https://scontent.xx.fbcdn.net/v/t1.6435-1/avatar_123.jpg",
  "profileId": "100004512378965",
  "profileName": "Maria Alvarez",
  "profileUrl": "https://www.facebook.com/maria.alvarez.921",
  "likesCount": 214,
  "commentsCount": 1,
  "threadingDepth": 0,
  "facebookId": "1234567890123456",
  "inputUrl": "https://www.facebook.com/humansofnewyork/posts/pfbid0BbKbkisExKGSKuhee9a7i86RwRuMKFC8NSkKStB7CsM3uXJuAAfZLrkcJMXxhH4Yl",
  "expansionToken": "ZXhwYW5zaW9uOnRva2VuOjEyMzQ1Njc4OTA=",
  "reply": [
    {
      "facebookUrl": "https://www.facebook.com/humansofnewyork/posts/pfbid0BbKbkisExKGSKuhee9a7i86RwRuMKFC8NSkKStB7CsM3uXJuAAfZLrkcJMXxhH4Yl",
      "commentUrl": "https://www.facebook.com/humansofnewyork/posts/pfbid0BbKbkisExKGSKuhee9a7i86RwRuMKFC8NSkKStB7CsM3uXJuAAfZLrkcJMXxhH4Yl?comment_id=1234567890123456&reply_comment_id=9988776655443322",
      "id": "Y29tbWVudDo5OTg4Nzc2NjU1NDQzMzIy",
      "feedbackId": "ZmVlZGJhY2s6OTk4ODc3NjY1NTQ0MzMyMg==",
      "date": "2026-06-02T15:41:09+00:00",
      "text": "Agreed, one of the best ones this year.",
      "profilePicture": null,
      "profileId": "100009887766554",
      "profileName": "Devon Clarke",
      "likesCount": 8,
      "commentsCount": 0,
      "reply": [],
      "threadingDepth": 1,
      "facebookId": "9988776655443322",
      "inputUrl": "https://www.facebook.com/humansofnewyork/posts/pfbid0BbKbkisExKGSKuhee9a7i86RwRuMKFC8NSkKStB7CsM3uXJuAAfZLrkcJMXxhH4Yl"
    }
  ],
  "url_key": "pfbid0BbKbkisExKGSKuhee9a7i86RwRuMKFC8NSkKStB7CsM3uXJuAAfZLrkcJMXxhH4Yl",
  "full_url": "https://www.facebook.com/humansofnewyork/posts/pfbid0BbKbkisExKGSKuhee9a7i86RwRuMKFC8NSkKStB7CsM3uXJuAAfZLrkcJMXxhH4Yl"
}

Field	Type	Description
facebookUrl	string	The input post/photo/reel URL this comment came from
commentUrl	string	Direct permalink to this specific comment
id	string	Facebook's internal comment node ID
feedbackId	string	GraphQL feedback ID used to paginate this comment's own replies
date	string (ISO 8601, UTC)	Comment timestamp
text	string	Comment body text
profilePicture	string | null	Commenter's profile picture URL
profileId	string	Commenter's numeric Facebook ID
profileName	string	Commenter's display name
profileUrl	string	Commenter's profile URL — present only when the author exposes one
likesCount	integer	Total reactions on the comment
commentsCount	integer	Number of direct replies Facebook reports for this comment
reply	array	Nested reply objects in this same shape, populated up to maxRepliesPerComment; empty on a reply object itself (only one reply level is expanded)
threadingDepth	integer	0 for a top-level comment, 1 for a reply
facebookId	string	Legacy numeric Facebook ID for the comment
inputUrl	string	The input URL that produced this row (mirrors facebookUrl)
expansionToken	string	Present only when the comment has an expandable reply thread
pageAdLibrary	object	{ is_business_page_active, id } — present only when the comment's author is a Page with Ad Library data
url_key	string	Short identifier parsed from the end of the input URL
full_url	string	The original input URL (duplicate of facebookUrl)

If a URL cannot be processed at all (bad permalink, blocked load, no comment pages returned), the Actor pushes a row with only url_key, full_url, error, and an empty comments array — and that row is pushed without the charged event, so failed URLs are never billed. To exclude these accounting rows from downstream processing, filter on the presence of text (real rows always have it) or the absence of error:

{ "url_key": "pfbid0Xyz...", "full_url": "https://www.facebook.com/PageName/posts/pfbid0Xyz...", "error": "Failed to extract required data from the post page", "comments": [] }

Schema stability and export options

Field names stay stable across runs — the row shape comes from a fixed conversion function, not from whatever markup Facebook happens to render, so a front-end redesign on Facebook's side does not rename your columns. Optional fields (profileUrl, expansionToken, pageAdLibrary) are simply omitted, never sent as empty strings, when Facebook doesn't expose them for a given comment. Results live in a standard Apify dataset and can be exported from the Output tab, or via the API, to JSON, CSV, Excel (XLSX), or XML.

💡 Facebook Comments Scraper Pay Per Events Strategy Guide

🎯 Strategy 1: Real-time enrichment pipeline

Trigger a run whenever a new post needs its comments attached to a lead, ticket, or CRM record: on the event (a new post detected by facebook-page-posts-scraper-with-lead-contact-list, a support ticket referencing a Facebook post, etc.), call this Actor with startUrls set to that one post URL and a modest commentSettings.maxComments. Poll the run's dataset via the Apify API, then append each row's text, profileName, and likesCount to the destination record. Because output is already typed JSON, no HTML parsing sits between the Actor and your CRM write.

🎯 Strategy 2: Scheduled monitoring and alerting

For ongoing tracking, schedule a recurring run against the same post or page URLs using Apify's Scheduler, with onlyCommentsNewerThan set to a rolling window ("1 day" for a daily job) so each run only pulls what's new since the last one. Diff the incoming id/facebookId values against what you stored from the previous run, and alert when commentsCount spikes on a specific comment or a new reply lands with an unusually high likesCount — the signal that a thread is taking off or turning negative.

🎯 Strategy 3: Bulk dataset build

For a research corpus or training set, paste every post URL you need into startUrls in one run — bulk paste is supported natively — and set commentSettings.maxComments and maxRepliesPerComment high enough for full coverage. Export the resulting dataset directly to CSV or JSON, or pull it via the API into a database. For very large URL lists, split them across multiple Actor runs within your Apify plan's concurrent-run allowance rather than relying on one oversized run.

Strategy comparison at a glance

Strategy	Best for	Run pattern	Output format
Real-time enrichment	CRM/lead enrichment on inbound events	Single on-demand run per event, via API or Console	JSON, polled via API
Scheduled monitoring	Ongoing sentiment/volume tracking on key posts	Recurring run via Apify Scheduler with onlyCommentsNewerThan	Dataset diffed run-over-run
Bulk dataset build	Research corpora, training data	One run with many startUrls	CSV/JSON export

🌴 Related Facebook Scrapers & Tools

Scraper Name	What it extracts
facebook-groups-comments-replies-scraper	Posts, comments, and nested replies from public Facebook Groups
facebook-posts-search-comments-scraper	Posts discovered by keyword search, plus each post's top comments
facebook-page-posts-scraper-with-lead-contact-list	A Page's posts and public contact/lead details
facebook-ads-library-scraper-advertiser-contact-finder	Ad Library ads and advertiser contact details
facebook-url-to-id-scraper-username-finder	Resolves Facebook URLs, usernames, and numeric IDs to each other
instagram-comments-scraper-with-engagement-analytics	Instagram post/reel comments and replies with engagement metrics
youtube-comments-scraper-with-reply-thread-expansion	YouTube video comments with fully expanded reply threads
reddit-comment-scraper-sentiment-lead-scoring	Reddit post comments with sentiment and lead scoring
LinkedIn-Company-Scraper-Post-Reactions-Comments-And-Engagement-Analytics	LinkedIn company page posts, reactions, and comments
facebook-user-search-scraper-with-lead-contact-enrichment	Facebook profiles and Pages discovered by name/niche, with contact enrichment

How to integrate Facebook Comments Scraper Pay Per Events with your stack

Facebook Comments Scraper Pay Per Events works with any language or tool that can call the Apify API — there's no dedicated SDK beyond the official Apify client libraries.

Python

from apify_client import ApifyClient
import csv

client = ApifyClient("<YOUR_APIFY_API_TOKEN>")

run_input = {
    "startUrls": [
        "https://www.facebook.com/humansofnewyork/posts/pfbid0BbKbkisExKGSKuhee9a7i86RwRuMKFC8NSkKStB7CsM3uXJuAAfZLrkcJMXxhH4Yl",
    ],
    "commentSettings": {
        "maxComments": 100,
        "maxRepliesPerComment": 5,
        "commentsSortOrder": "newest",
    },
    "onlyCommentsNewerThan": "2 weeks",
}

run = client.actor("YOUR_USERNAME/facebook-comments-scraper-pay-per-events").call(run_input=run_input)

rows = []
for item in client.dataset(run["defaultDatasetId"]).iterate_items():
    if item.get("error"):
        continue  # uncharged accounting row — skip
    rows.append({
        "post": item.get("full_url"),
        "author": item.get("profileName"),
        "text": item.get("text"),
        "likes": item.get("likesCount"),
        "date": item.get("date"),
        "replies": len(item.get("reply", [])),
    })

with open("facebook_comments.csv", "w", newline="", encoding="utf-8") as f:
    writer = csv.DictWriter(f, fieldnames=["post", "author", "text", "likes", "date", "replies"])
    writer.writeheader()
    writer.writerows(rows)

print(f"Saved {len(rows)} comments to facebook_comments.csv")

Node.js

import { ApifyClient } from 'apify-client';

const client = new ApifyClient({ token: '<YOUR_APIFY_API_TOKEN>' });

const input = {
    startUrls: ['https://www.facebook.com/humansofnewyork/posts/pfbid0BbKbkisExKGSKuhee9a7i86RwRuMKFC8NSkKStB7CsM3uXJuAAfZLrkcJMXxhH4Yl'],
    commentSettings: { maxComments: 50, maxRepliesPerComment: 2, commentsSortOrder: 'most_relevant' },
};

const run = await client.actor('YOUR_USERNAME/facebook-comments-scraper-pay-per-events').call(input);
const { items } = await client.dataset(run.defaultDatasetId).listItems();

const comments = items.filter((item) => !item.error);
console.log(`Fetched ${comments.length} comments`);
comments.forEach((c) => {
    console.log(`${c.profileName}: ${c.text} (${c.likesCount} likes, ${(c.reply || []).length} replies)`);
});

Async and scheduled pipelines

Every run is started on demand through the Apify Console or API and returns a dataset ID you can poll for completion — there is no actor-specific webhook or push mechanism in the source. For recurring jobs, use Apify's platform-level Scheduler to fire this Actor on a cron interval, and either poll the run status via the API or configure an Apify webhook on the run's completion event so your system is notified instead of polling continuously.

🎯 Who Needs Facebook Comments Scraper Pay Per Events? (Use Cases & Industries)

🏢 Social listening, brand monitoring and moderation teams

A community manager pulls comments on a brand's latest campaign post with commentsSortOrder: newest, scans text and likesCount for spikes in complaints or praise, and escalates any comment that needs a public reply — using profileName and commentUrl to respond directly on the original thread.

📊 AI/ML teams building training data and RAG pipelines

A data team indexes text, profileName, and date from thousands of comments into a vector store for a customer-sentiment RAG assistant, using likesCount as a lightweight relevance signal when ranking retrieved snippets.

📱 Competitive intelligence and marketing analysts

An analyst tracks comment volume and likesCount on a competitor's product-launch posts week over week, using onlyCommentsNewerThan to isolate what changed since the last check and spotting which specific claims drew the most engaged responses.

🔬 Researchers

Academic and social-science researchers collect public comment threads on news or public-health posts to study discourse patterns and public sentiment, scoped entirely to publicly reachable permalinks — no login-gated or private-group content is ever included.

🎥 Product and SaaS builders

A SaaS team building a social-listening dashboard schedules this Actor against client-owned Pages, maps text/profileName/likesCount into their own normalized schema, and layers sentiment scoring or alerting on top without maintaining a Facebook HTML parser themselves.

Is it legal to scrape Facebook?

Scraping publicly accessible Facebook posts and comments — content visible without logging in — is generally treated differently under U.S. law than bypassing a login wall or authentication barrier. In hiQ Labs, Inc. v. LinkedIn Corp. (9th Circuit Court of Appeals, 2019), the court held that scraping data that is not access-restricted does not constitute "unauthorized access" under the Computer Fraud and Abuse Act — a widely cited general principle for public-data scraping in the U.S., though it did not involve Facebook specifically and its reasoning may not extend to every jurisdiction or every category of data.

Separately, Meta's own Terms of Service prohibit automated data collection without permission; violating those terms is a civil contract matter — grounds for an account or IP ban, or a breach-of-contract claim — not a criminal offense. Because this Actor's output includes personal data about real individuals (profileName, profileId, profileUrl, profilePicture), storing or processing that data may also trigger obligations under regimes like the GDPR (EU/EEA) or CCPA (California) depending on who you are and where your users are.

Facebook Comments Scraper Pay Per Events returns only publicly accessible data. What you do with that data is your responsibility — consult legal counsel for commercial applications involving personal data.

❓ Frequently asked questions

Does Facebook Comments Scraper Pay Per Events work without a Facebook account?

Yes. The Actor loads the public post, photo, or reel permalink logged-out and paginates Facebook's own comment API directly — no login, session cookie, or Graph API access token is read from or required by the input schema.

How does it handle Facebook's anti-scraping measures?

It always routes every request through Apify's residential proxy pool, regardless of what's set in proxyConfiguration, and retries failed HTTP requests up to three times with exponential backoff before giving up on a given fetch.

Can I run it at scale without getting blocked?

startUrls accepts a bulk list of post URLs processed in a single run, and each run's size is governed by the maxComments/maxRepliesPerComment values you set — no uptime or success-rate figure is published for this Actor, so plan capacity by your own testing rather than an assumed number.

How fresh is the data it returns?

Live. Each run issues fresh GraphQL requests to Facebook at execution time; nothing is served from a cache.

Which fields work best for AI training and RAG indexing?

text is the primary field for both — it's the actual comment content. For RAG, pair it with profileName, date, and facebookUrl for citation context. For training data, likesCount, commentsCount, and threadingDepth (top-level vs. reply) give consistent, typed structure across every record with no normalization required.

Does this Actor collect personal data, and who's responsible for it?

Yes — profileName, profileId, profileUrl, and profilePicture identify real commenters. The Actor only returns what Facebook already displays publicly; the lawful basis for storing, processing, or acting on that data sits with you as the operator, not with the Actor.

Does Facebook Comments Scraper Pay Per Events work with Claude, ChatGPT, and other AI agent tools?

Yes, as an HTTP endpoint — any agent framework that can call the Apify API (directly, or through the official apify-client SDK) can start a run and read back typed JSON rows with no parsing step, which is what an LLM context window needs.

What happens when a URL fails to load?

The Actor pushes a row containing only url_key, full_url, error, and an empty comments array for that URL, and continues to the next one instead of failing the whole run. This accounting row is pushed without the charged event, so a failed URL is never billed.

How does the date filter (onlyCommentsNewerThan) actually work?

It's applied after comments are fetched, not as a query-time filter against Facebook — so for a narrow recent window, set maxComments generously to make sure enough of the thread is pulled before filtering trims it down.

What's the difference between maxComments and maxRepliesPerComment?

maxComments caps top-level comments fetched per URL. maxRepliesPerComment separately caps how many nested replies are fetched for each of those top-level comments — replies never count against the maxComments limit.

ℹ️ Disclaimer

Facebook Comments Scraper Pay Per Events extracts only publicly available data from Facebook. This tool is intended for lawful use cases only. Users are responsible for complying with Facebook's terms of service and applicable data protection laws in their jurisdiction.