Facebook & Meta Ad Library Scraper: by website, with EU reach

Type a website. Get every Facebook and Instagram ad that brand is running right now.

No page IDs or Ad Library links needed: give it gymshark.com and it finds Gymshark's own Facebook pages for you, leaving out look-alike pages with the same name.

👀 See it on real brands first: the free European Ad Watch shows what 29 European brands are advertising this week, made with this tool and updated every Monday.

What you get

For every ad

• ⏱️ How long it has run: start date and days running
• 📑 How many copies of it run: brands copy the ads they put money behind
• 🔗 The ad itself: text, headline, button, landing page, image and video
• 🇪🇺 Who saw it in the EU (optional): people reached by country, age and gender, plus the targeting and who paid

For every brand, one summary row

• 📊 How many ads Meta shows for them, and how many different creatives that is
• 🆕 How many of the ads read started this week and this month
• 🏆 The 5 ads they back hardest
• 🔄 What's new and what stopped since your last run, if you run it on a schedule

How to use it

1. Add websites (or Facebook pages, page IDs or brand names), one per line.
2. Choose Every running ad, or One row per advertiser to check many companies cheaply.
3. Click Start. Download the result as Excel, CSV or JSON, or send it to Google Sheets, Make, n8n, Zapier or your AI agent.

Tip: schedule it weekly, and each run tells you which of your competitors' ads are new.

Pricing

You pay only for results:

What	Price
Each brand that is advertising	$0.02
Each ad	$0.0015 ($1.50 per 1,000 ads)
EU reach for an ad (optional)	$0.003
Saved copy of an ad's image (optional)	$0.001

• Brands that aren't advertising, and websites that couldn't be read, cost nothing.
• A run never goes over the spending limit you set.

Examples: 10 competitors with 50 ads each costs $0.95. Checking whether 1,000 companies advertise costs at most $20, and nothing for the ones that don't.

Example: one ad

{
  "pageName": "HelloFresh",
  "headline": "Spare bis zu 120 €",
  "startDate": "2026-09-17",
  "daysRunning": 22,
  "copiesRunning": 3,
  "platforms": ["facebook", "instagram", "threads"],
  "format": "video",
  "cta": "Order now",
  "landingDomain": "hellofresh.de",
  "isNewSinceLastRun": false,
  "euAudience": {
    "reach": 2270985,
    "countries": [{ "country": "DE", "reach": 2297479 }],
    "genderSplit": { "female": 0.71, "male": 0.25, "unknown": 0.03 },
    "topAgeGroup": "25-34",
    "payer": "HelloFresh SE"
  },
  "adLibraryUrl": "https://www.facebook.com/ads/library/?id=1201251448800306"
}

Example: one brand

{
  "input": "elixia.fi",
  "status": "advertising",
  "matchedBy": "Linked from elixia.fi",
  "activeAds": 49,
  "libraryTotal": 49,
  "distinctCreatives": 12,
  "startedLast7Days": 11,
  "longestRunningDays": 49,
  "mostBackedAds": [{ "copiesRunning": 12, "daysRunning": 49, "headline": "Aloita tästä 💙" }],
  "newSinceLastRun": 0,
  "stoppedSinceLastRun": 4
}

In Every running ad mode, these brand rows are saved in the run's storage under OVERVIEW.

Good to know

• Running ads only. That's what Meta's library lists. With a schedule, stopped ads are counted from your second run.
• No spend or clicks. No ad library publishes them, and this Actor doesn't guess them. Reach is published only for ads shown in the EU.
• "Copies running" is a signal, not proof that an ad is profitable. It shows where a brand puts its weight.
• Up to 500 ads per brand, in the order Meta lists them, which is not by date (set Max ads per advertiser; 100 when left empty). Read them all for exact counts of new ads. The summary shows how many Meta lists next to how many were read, so nothing is silently missing.
• EU reach by country can add up to a bit more than the total: Meta counts a person once in the total, but in every country they were reached in.
• Saved images are what Meta publishes. For video ads that's the preview frame, sometimes a small one.

For AI agents and automations

• AI agents (Claude, ChatGPT, Cursor) can run it through Apify's MCP server. Ask for One row per advertiser first: one short row per brand fits an agent's context.
• Clay, n8n, Make, Zapier: run it per company with the website as input. Field names stay stable; new fields are only added.
• It keeps three small stores in your account so later runs are faster and can say what changed (reachlens-meta-lookups, -history, -images). Delete them to start fresh.

Also by Reachlens: Google Ads Transparency Scraper, the same idea for Google Search, YouTube and Display ads, including the words inside image ads.