Facebook/Meta Ads Library Scraper API - Winning Ads, Scam Check MCP server

Facebook/Meta Ads Library Scraper API - Winning Ads, Scam Check

Scrape the Facebook Ads Library, which Meta now calls the Meta Ad Library (ads on Facebook, Instagram, Messenger, Threads and Audience Network), with no login, no cookies and no Meta API token. Type a keyword or paste a Facebook page, and get every matching ad as a row: the advertiser, how many days the ad has run, how many versions it has, the text, headline and button, the landing page and its domain, image and video links, platforms, and the reach, spend and EU audience data Meta publishes. On top of that: an advertiser list (businesses advertising on a keyword right now, one row each) and a brand scam check (ads that use a brand's name from pages that are not the brand's own). $0.70 per 1,000 ads, no start fee. Use it from the Apify Console, the API, Python, JavaScript, n8n, Make, Zapier or an AI agent through the Apify MCP server.

What does it do?

It reads the public Meta Ad Library, the same one at facebook.com/ads/library, and returns what you would otherwise copy by hand, for as many keywords or advertisers as you give it:

• Ads for keywords (ads): every ad matching your words in a country or worldwide.
• Ads of advertisers (pages): all ads of the Facebook pages you list, by link, page id or just the name.
• Advertiser list (advertisers): the businesses advertising on your keywords now, with how many ads, their longest-running ad and the websites they send people to.
• Brand scam check (impersonation): ads that name a brand but come from other pages, each with a risk level and the signals behind it (look-alike page name, look-alike landing domain, a new ad from a small page).

Why do people use the Ad Library data?

Answer: to see what already works in someone else's ads. Meta shows every running ad but no results, so the best free signal is time: an ad that has run for 30 to 90 days is almost always paying for itself, because nobody keeps paying for an ad that loses money.

• Media buyers and DTC brands find competitors' long-running ads and copy the hook, offer and format instead of paying to test from zero. daysRunning and versions are here for that: several live versions of one ad are an A/B test, and the survivors are the winners.
• Creative strategists and agencies keep a weekly watchlist of 10 to 20 brands and build swipe files; the landing domain shows the funnel behind each ad.
• Agencies prospecting use the advertiser list: a business spending on ads now has a budget, so it is a warm lead.
• E-commerce and dropshipping spot products pushed with many long-running ads.
• Developers build ad-spy tools, n8n and Make flows (scrape, then have GPT or Gemini pull out hooks and write new variations) and AI-agent tools on top of it, without maintaining a scraper.
• Brand protection and security teams run the brand scam check. Meta itself estimated about 15 billion scam ads are shown every day and expanded its brand-impersonation detection in March 2026; this finds the ones using your name.

How is it different from other Facebook Ads Library scrapers?

Shortened. The full README is on GitHub.