Back to guides
Rows of Facebook ads exported from the Meta Ad Library into a spreadsheet
Meta Ad LibraryAutomationHow-To

Facebook Ads Library Scraper: 5 Ways to Export Ads

The Meta Ad Library has no export button. If you want competitors' ads in a spreadsheet, a database or an AI tool, you have five options, and the one most people try first (the official API) is the one least likely to work for ordinary commercial ads.

This guide compares all five, with real prices for the most-used scrapers checked on 10 November 2026, and the mistakes that make a scraped dataset useless after a week.

Disclosure: we make Seercraft, and one of the Apify scrapers below is built by Anas Nadeem, who also builds Seercraft. We've listed it with its real price, which is the highest of the four.

Why can't you just use the Facebook Ad Library API?

Because it only returns most ads if they were delivered in the EU. Meta's API returns ads that reached any EU location, and outside the EU only social issue, electoral and political ads. A US or Indian ecommerce brand's ads simply don't come back.

If you do qualify (EU ads, or political and issue ads), the API is the cleanest route. Every call needs a country, search terms are capped at 100 characters, you can pass up to ten Page IDs at once, and rate limits apply. We cover it in depth in Facebook Ad Library API: What It Can and Can't Return.

How do Apify scrapers compare?

Apify is a marketplace of hosted scrapers ("actors") that read the public Ad Library pages and return rows. You give them a search URL or Page, they return ads as JSON, CSV or Excel. Most charge per ad. These are the four most-used ones we found for Meta's library, priced on Apify's free tier:

ActorUsersPrice on the free tierNotes
Apify's own Facebook Ads Scraper51,880$5.80 per 1,000 adsDrops to $1.00 per 1,000 on Apify's top tier. Optional e-commerce add-on
igolaizola's Facebook (Meta) Ads Library Scraper4,604$0.75 per 1,000 ads, plus $0.0075 per runAd detail fetches are charged separately
automation-lab's Facebook Ads Library Scraper5,723$0.58 per 1,000 ads, plus $0.005 per runCheapest per ad on the free tier
Meta Ad Library Scraper (whoareyouanas, built by Seercraft's maker)3,780$10 per 1,000 ads, flatHighest price of the four. Flat rate on every tier

User counts and prices come from each actor's public Apify listing. Higher Apify tiers cut most per-ad prices, so check the listing for your plan.

What to check before you pick one:

  • Which fields come back. Most return the Library ID, Page, start date, platforms, ad copy, link, call to action and media URLs. Some add Page follower counts or a reach figure for EU ads.
  • Media URLs expire. Image and video links point to Meta's CDN and stop working after a while. If you want the creative, download it in the same run.
  • Cost per useful ad. A brand with 5,800 active ads costs 5,800 rows, even though most are copies of a few ideas. Filter by start date or media type in the search URL first.

Do Chrome extensions work?

Several Chrome extensions export the Ad Library page you're looking at to CSV, or save single ads into a board. They're fine for a few dozen ads. They break when Meta changes the page layout, and they only capture what you've scrolled. Read the permissions before you install one: an extension that can "read and change all your data on all websites" can read every site you visit.

Seercraft's own Chrome extension is the save-one-ad kind: it adds a Save to Seercraft button to ads in the library, and it keeps the media so the ad still plays after it stops running.

Should you build your own scraper?

Only if scraping is your product. A headless browser (Playwright or Puppeteer) can load an Ad Library search URL and read the ad cards, and for a one-off pull it works. Keeping it working is the hard part: Meta changes markup and rate limits without notice, large Pages cut off results (the library itself says "We're unable to display every ad" for brands like Quince), and you need proxies at any volume.

Also read Meta's terms on automated data collection before you build anything that runs on a schedule. Hosted scrapers carry the same question, so decide what you're comfortable with.

What do people get wrong with scraped ad data?

  1. Overwriting instead of appending. Commercial ads disappear from the library when they stop. If each run replaces the last, you lose the history, which is the useful part. Key every row on the Library ID and keep first_seen and last_seen.
  2. Counting copies as ideas. Big advertisers run dozens of near-identical versions of one ad. Group by creative and text before you count anything, or "200 new ads" turns out to be four ideas.
  3. Ignoring start dates. A cluster of ads with the same start date is usually one launch. A creative still running after months is the best public hint that it works. See how to find the longest-running Facebook ads.
  4. Trusting "spend" columns. Meta doesn't publish spend for commercial ads. Any spend or budget field from a scraper is an estimate.

When should you skip scraping entirely?

When what you actually want is to watch a handful of competitors every week. Then a dataset is a means, not the goal, and maintaining one is a side project.

That's the job Seercraft does: you add the brands you compete with and it reads their ads in the public Ad Library every day, keeps each ad after it stops, groups copies into one idea with a version count, and every Monday Seer, the built-in analyst, sends three tests to run with links to the ads behind them. There's no CSV export. If you want the data inside Claude, ChatGPT or Cursor instead, Seercraft's MCP server lets you ask about any tracked brand with live data and links.

If you need raw rows for your own analysis, an Apify actor is the faster route. If you need the weekly answer, use a tracker. Every Seercraft plan starts with a 7-day free trial; see pricing.

Sources

Published

Stop scrolling the Ad Library at 1am.

See what's working in about a minute. Start with a 7-day free trial. Cancel anytime.