Meta Ads Library Scraper

What does Meta Ads Library Scraper do?

Meta Ads Library Scraper finds the Facebook and Instagram ads that a website is running. Enter one or more domains (for example nike.com) and the Actor returns the matching ads from the public Meta Ads Library as clean, structured data: ad text, headline, landing page, start and end dates, platforms, images, videos and product cards.

It matches the exact domain, so you get the ads for that website, not a loose keyword search. Because it runs on Apify, you also get API access, scheduling, webhooks, integrations and an MCP tool for AI agents.

Why use Meta Ads Library Scraper?

• Competitor research: see what your competitors advertise, which products they push and how their ads change.
• Lead generation and sales: find companies that actively spend on Meta ads.
• Agency reporting: pull a client's live ads, or a whole list of brands, in one run.
• AI agents: answer "what ads is example.com running in Germany?" through the Apify MCP server.

Why the results are more precise

The Actor searches Meta's Ads Library for the exact domain phrase. Scrapers that use a broader keyword search often return ads from other advertisers that only mention similar words. With exact-domain matching, the result list stays close to the ads that belong to the website you asked about. Meta shows an approximate total (for example "~280 results") and the Actor reports it in the run summary so you can compare.

Which ads do you get first?

Ads are sorted by total impressions, highest first (the same ranking as Meta's own "most impressions" sort). If a domain has 280 ads and you set maxAdsPerDomain to 50, you get the 50 ads with the most impressions, not the 50 newest.

How to scrape Meta ads by domain

1. Open the Input tab and enter one or more domains, for example nike.com.
2. Optionally choose a country, ad status, platforms and media type.
3. Click Start. A single domain with up to 30 ads usually finishes within seconds.
4. Open the Output tab and download the ads as JSON, CSV, Excel or HTML, or read them through the API.

You can paste domains in any form. The Actor removes https://, www., paths and query strings, and combines duplicates, so each unique domain is checked once and charged once, even if it appears several times in your list.

Only website domains are searched. Entries that are not valid domains, such as a brand name or a keyword (nike, test), are skipped and not charged, so you never pay for a result that is not about a website.

Input

Field	Type	Default	Description
domains	array of strings	required	1 to 100 domains, such as ["nike.com"]. Full URLs are cleaned to the domain. Use domains, not brand names or keywords; other entries are skipped and not charged.
maxAdsPerDomain	integer	30	1 to 500 ads per domain, highest total impressions first.
country	string	ALL	2-letter country code (for example DE, US) or ALL.
activeStatus	string	active	active for ads running now, all to include past ads.
publisherPlatforms	array	all	Any of facebook, instagram, audience_network, messenger, whatsapp, threads.
mediaType	string	all	all, image, meme, image_and_meme, video, none.

No proxy setup is needed. For larger requests the Actor handles it automatically, and the cost is included in the price.

Example input

{
  "domains": ["nike.com", "adidas.com"],
  "maxAdsPerDomain": 50,
  "country": "US",
  "activeStatus": "active"
}

Output

The dataset has one row per ad, and every domain you enter appears in the results:

• Ad rows have hasAds: true and all ad fields.
• A domain with no ads gets one row with hasAds: false. Only domain, hasAds and scrapedAt are filled, all other values are null.
• A domain that could not be checked (for example when Meta did not respond) gets one row with hasAds: null. It is not charged.
• An entry that is not a valid domain also gets one row with hasAds: null and is not charged. The summary marks it invalid.
• A domain that was skipped because the run reached its time limit or your maximum cost gets one row with hasAds: null and is not charged. The summary marks it skipped.

You can download the dataset in various formats such as JSON, HTML, CSV or Excel.

{
  "domain": "alternate.de",
  "hasAds": true,
  "adArchiveId": "1793908174954881",
  "adLibraryUrl": "https://www.facebook.com/ads/library/?id=1793908174954881",
  "pageName": "ALTERNATE",
  "headline": "ASUS PRO WS W790-ACE, Mainboard",
  "bodyText": "Warum noch zögern? Dein Setup verdient mehr! ...",
  "landingPageUrl": "https://www.alternate.de/listing.xhtml?b=1220&bc=11",
  "ctaText": "Shop now",
  "isActive": true,
  "startDate": "2026-07-27T07:00:00.000Z",
  "platforms": ["FACEBOOK", "INSTAGRAM", "AUDIENCE_NETWORK", "THREADS"],
  "displayFormat": "DPA",
  "cards": [
    { "title": "ASUS PRO WS W790-ACE, Mainboard", "linkUrl": "https://www.alternate.de/...", "imageUrl": "https://..." }
  ]
}

Data fields

Field	Description
domain	The domain from your input this row belongs to
hasAds	true on ad rows, false for a domain without ads, null if it could not be checked or is not a valid domain
adArchiveId, adLibraryUrl	Meta's ad ID and a link to the ad in the Ads Library
pageName, pageId, pageProfileUrl, pageLikeCount	The advertiser's Facebook page
headline, bodyText, linkDescription	Ad headline and texts
landingPageUrl, ctaText	Where the ad sends people and its button text
isActive, startDate, endDate	Status and dates (ISO 8601, UTC)
platforms, displayFormat	Where the ad runs and its format (IMAGE, VIDEO, CAROUSEL, DPA)
imageUrls, videoUrls	All creative URLs of the ad
cards	Slides of carousel and catalog ads: product title, link and image

Every field is described in the Actor's dataset schema.

Run summary

Each run also saves a SUMMARY record in the key-value store with one entry per domain: its status (ok, no_ads, partial, skipped, invalid, error, no_ads_read), the number of ads returned, Meta's approximate total, and a note. It is not part of the dataset and not charged. Check it when a domain returns fewer ads than you expected.

If Meta's Ads Library is not responding and no domain could be checked, the run fails with a clear message, so you (or your agent) know to try again later. That is different from a domain that simply has no ads. If none of the entries is a valid domain, the run also fails right away with a message saying so.

How much does it cost to scrape Meta ads?

You pay per event, and platform and proxy costs are included, so there are no extra usage charges and no start fee:

• Domain checked: $0.005 per unique domain, on every plan. It applies whether the domain has ads or not. Domains that could not be checked are not charged.
• Ad scraped: a price per ad that depends on your Apify plan. It applies to every ad saved.

Apify plan	Domain checked	Ad scraped	Per 1,000 ads
Free	$0.005	$0.001	$1.00
Starter	$0.005	$0.0006	$0.60
Scale	$0.005	$0.00055	$0.55
Business	$0.005	$0.0005	$0.50

Examples for one domain:

Result	Free	Starter	Scale	Business
No ads found	$0.005	$0.005	$0.005	$0.005
30 ads	$0.035	$0.023	$0.0215	$0.020
100 ads	$0.105	$0.065	$0.060	$0.055

You can set a maximum cost per run in the Apify Console. The Actor stops when it is reached and marks the affected domains in the summary.

Tips for faster and cheaper runs

• Keep maxAdsPerDomain at 30 when you only need a sample of the most-seen ads. Up to about 30 ads is the fastest.
• Set a country to match the market you care about. Results are filtered by the ads shown there.
• Use activeStatus: active to skip old ads.
• Group domains in one run instead of starting many runs. Several domains are processed in parallel.
• Screening a long list? If you only need to know which domains advertise at all, my companion Actor Meta Ads Checker is a cheaper yes/no check. Then scrape just the domains with ads here.

Use Meta Ads Library Scraper with AI agents (MCP)

Add this Actor as a tool to the Apify MCP server:

https://mcp.apify.com/?tools=m3web/meta-ads-library-scraper

Example prompts for your agent:

• "What ads is nike.com running on Instagram in the US right now?"
• "List the products adidas.com is advertising in Germany."
• "Compare how many active ads alternate.de and mediamarkt.de run."

The Actor accepts plain domains or URLs and checks each unique domain once. It returns flat, documented fields, and every domain gets at least one row, so an agent can tell hasAds: true, false (no ads) and null (could not be checked or not a valid domain) apart without guessing. The per-domain summary adds the exact status and a note.

FAQ, disclaimers and support

Why did I get fewer ads than maxAdsPerDomain? The domain may have fewer ads, or not all of them could be loaded in this run. Check the SUMMARY record for the status and a note. Meta's total is approximate and can differ slightly from the number of rows.

Which ads come first if I set a limit? The ads with the most total impressions. Sorting by date is not available yet.

Can I search by brand name, keyword or Facebook page? No. The Actor searches for the exact website domain, so a brand name or keyword is skipped and not charged. Use the brand's website domain.

Can I get impressions, spend or audience data? impressionsText is filled only when Meta publishes it, mainly for political and social-issue ads. This Actor does not collect audience breakdowns.

Do image and video links keep working? They are signed by Meta and can stop working after a few days. Download files you need to keep.

Is it legal to scrape the Meta Ads Library? The Actor reads publicly available pages of Meta's Ads Library. You are responsible for using the data in line with applicable laws, Meta's terms and data protection rules. Do not use it to collect personal data.

Found a problem or need another field? Open an issue in the Issues tab of this Actor. Custom solutions are available on request.