Facebook & Meta Ads Library Scraper + Tracker

Use this Facebook Ads Library scraper to collect current Facebook and Instagram ads from the Meta Ad Library, or reuse the same search as a change feed for competitor monitoring. It also works as a Meta ads scraper for one-time exports: add a keyword, Facebook page, or prepared Ad Library URL and receive normalized ad copy, creative URLs, calls to action, dates, advertiser context, destinations, and explicit coverage status.

For the easiest first run, choose Baseline, keep the prefilled nike search and 15-ad limit, then click Start.

🚀 Start in 60 seconds

1. Keep Baseline — current ads.
2. Enter a keyword, a Facebook page URL, or an Ad Library search URL.
3. Keep 15 ads per target for a small first test.
4. Click Start and open the Dataset when the run finishes.

{
  "productLevel": "BASELINE",
  "searchTerms": ["nike"],
  "country": "US",
  "activeStatus": "ACTIVE",
  "maxAdsPerTarget": 15
}

You can combine input lists, but one run accepts at most 25 total targets across searchTerms, pageUrls, and startUrls. The Console starts with a buyer-friendly 15-result prefill; API and Task inputs that omit maxAdsPerTarget keep the compatible default of 100.

🎯 Choose the right workflow

Workflow	Use it when	First run	Later run with the same scope	Result billing
Baseline	You need the ads visible now	Returns current matching ads	Returns a fresh current snapshot	One baseline_ad_result per delivered row
Changes	You want repeat monitoring	Creates state and returns current ads as NEW	Returns the current snapshot labelled NEW, UPDATED, UNCHANGED, plus any safely proven ENDED ads	One changes_ad_result per delivered row

You do not need to run Baseline before Changes. The first successful Changes run creates its own comparison state.

For meaningful comparisons, repeat the same:

• target or prepared Ad Library URL;
• country, ad category, status, format, and date filters;
• maxAdsPerTarget value.

Run the same monitoring scope sequentially. Changing the scope creates a different comparison, and overlapping runs of the same scope should be avoided.

📦 What you get

Each row is a normalized ad record. Fields are returned when Meta exposes them for that ad and category.

Field group	Examples	Availability
Ad identity	adId, pageId, pageName, adLibraryUrl	Core identity fields
Dates and state	startDate, endDate, state, isActive	Source-dependent
Copy and CTA	bodies, cardTitles, cardDescriptions, ctaTexts	When present in the ad
Creative media	imageUrls, videoUrls, snapshotUrls, displayFormat	Depends on creative format
Delivery context	publisherPlatforms, country, targetedOrReachedCountries	When Meta publishes it
Destination evidence	observedLinkUrls, observedDestinationUrl, resolution fields	When a destination is exposed and can be evaluated
Political-ad disclosures	reachEstimate, spend, currency, impression bounds	Political and issue ads only, when published
Monitoring	changeStatus, changedFields, changeEventId, observedAt	Changes workflow
Coverage	coverageStatus, sourceExhausted	Explains whether collection was complete or limited
Stable automation and provenance	stableId, sourceClass, sourceName, sourceUrl	Deduplication key, official-source identity, and canonical public source URL

Representative Changes row:

{
  "adId": "388120997678322",
  "pageId": "1497383683837785",
  "pageName": "adidas",
  "state": "INACTIVE",
  "bodies": ["Discover the latest collection."],
  "ctaTexts": ["Shop now"],
  "imageUrls": ["https://scontent.example.fbcdn.net/ad-creative.jpg"],
  "publisherPlatforms": ["FACEBOOK", "INSTAGRAM"],
  "country": "US",
  "changeStatus": "NEW",
  "coverageStatus": "COMPLETE",
  "observedAt": "2026-08-20T11:45:21.357Z"
}

The Dataset includes focused views for quick review:

• Buyer overview — the most useful advertiser, creative, destination, change, and coverage fields;
• New, updated and ended ads — change-first monitoring view;
• Creative research swipe file — copy, CTA, media, and destination research;
• All fields — complete normalized export for API, CSV, or downstream analysis.

Blank or absent optional fields mean the source did not expose that value for that ad. They are not filled with guesses.

🎛️ Input guide

Input	What it controls	Practical advice
productLevel	Current snapshot or repeat comparison	Start with Baseline; use Changes for scheduled monitoring
searchTerms	Keyword and brand searches	One term per line; nike is a safe test
pageUrls	Specific Facebook advertisers	Numeric-page-ID or Ad Library page URLs are the most reliable
startUrls	Searches prepared in Meta Ad Library	Filters embedded in the URL take precedence for that URL
maxAdsPerTarget	Result ceiling for each target	Console prefill is 15; omitted API/Task input defaults to 100; reaching it produces CAPPED coverage
adType	Commercial or political/issue ads	Political ads can expose reach and spend fields
country	Ad delivery country	ALL is supported only for political/issue ads; commercial ALL runs use and report US
activeStatus	Active, inactive, or all ads	Filter before collection to avoid irrelevant delivered rows
mediaType	Any, image, video, or meme creatives	Leave ALL unless the format itself is your research question
dateFrom, dateTo	First-shown date range	Optional; use YYYY-MM-DD
Advanced limits	Pagination, network, and timeout ceilings	Safe defaults protect cost; lower limits can make coverage partial

For a low-cost first run, set maxAdsPerTarget explicitly to 15 in API calls. Otherwise the compatible default is 100.

blockMediaDownloads should normally stay on. It reduces bandwidth by skipping media-file downloads during collection while still returning the published image and video URLs.

💰 Pricing

This Actor uses Pay per event. The current active price is:

Total = $0.02 run start + delivered rows × the result price for the selected workflow

The actor_start event is configured with isOneTimeEvent: true: it is a flat $0.02 once per successful run, not $20 per run. It is preflighted before collection and charged only after the Actor has a usable source collection or a positively verified empty result. A blocked, failed, or no-collection attempt charges zero start events.

Apify plan	Baseline per 1,000 rows	Changes per 1,000 rows	1,000-row Baseline total	1,000-row Changes total
Free	$0.700	$1.100	$0.720	$1.120
Bronze	$0.630	$1.045	$0.650	$1.065
Silver	$0.560	$0.990	$0.580	$1.010
Gold	$0.490	$0.935	$0.510	$0.955
Platinum	$0.455	$0.935	$0.475	$0.955
Diamond	$0.420	$0.935	$0.440	$0.955

Examples on the Free plan:

Example run	Calculation	Maximum event charge
15 Baseline rows	$0.02 + 15 × $0.0007	$0.0305
100 Baseline rows	$0.02 + 100 × $0.0007	$0.0900
100 Changes rows	$0.02 + 100 × $0.0011	$0.1300
1,000 Changes rows across multiple targets	$0.02 + 1,000 × $0.0011	$1.1200

At 1,000 delivered Baseline rows, Gold reduces the row price by 30% and the complete event charge by 29.2% after the flat start. For 1,000 Changes rows, Gold reduces the row price by 15% and the complete event charge by 14.7%. Platinum and Diamond keep the same Changes price as Gold; their additional discount applies to Baseline rows.

In Changes mode, every delivered NEW, UPDATED, UNCHANGED, and safely proven ENDED row uses the Changes price because each repeat run performs a fresh live scan. State operations, summaries, suppressed rows, and failed delivery attempts are not separate result events.

The examples show event charges at the listed rates. Your Apify account still applies its own platform billing rules shown at run time.

✅ Coverage you can trust

The Actor reports why a target stopped instead of presenting every short result as complete.

Status	Meaning
COMPLETE	The source was exhausted within the selected limits
CAPPED	maxAdsPerTarget was reached before source exhaustion
PARTIAL	A timeout, budget, pagination ceiling, or recoverable source problem prevented complete collection
EMPTY	No matching ad was found; sourceExhausted shows whether that empty result was fully proven
CHALLENGED	The source blocked or challenged collection

ENDED is conservative. It is emitted only when the previous comparison state and the current scan provide complete-enough evidence. A CAPPED, PARTIAL, or challenged run cannot convert a missing ad into ENDED.

If collection cannot meet that contract, the run reports the limitation or fails closed instead of silently claiming complete coverage.

🔌 API and automation

Use the same stable input keys from the Console through the Apify API:

const run = await client.actor('luminar/meta-ads-library-creative-changefeed').call({
  productLevel: 'CHANGES',
  searchTerms: ['nike'],
  country: 'US',
  activeStatus: 'ACTIVE',
  maxAdsPerTarget: 100,
});

const { items } = await client.dataset(run.defaultDatasetId).listItems();

After one successful manual run, save the input as your own private Apify Task and add a Schedule. Webhooks, Make, Zapier, n8n, or your own code can then filter NEW, UPDATED, and ENDED rows for alerts while retaining UNCHANGED rows as context.

For Google campaign research, use the companion Google Ads Transparency Center Scraper & Tracker.

❓ FAQ and support

Do I need to run Baseline before Changes?

No. The first Changes run establishes state and returns current ads as NEW. Repeat the same scope to compare later runs.

Why did a repeat Changes run return UNCHANGED rows?

Meta Changes is a labelled live snapshot. It returns the current monitored rows on every run so an automation has complete current context; those delivered rows use the Changes result price.

Why are copy, spend, reach, or media fields missing on some ads?

Meta exposes different fields by ad type and creative format. Spend and reach disclosures are primarily available for political and issue ads. Missing values remain absent instead of being inferred.

Can I schedule competitor monitoring?

Yes. First confirm the input manually, then save it as a private Task and attach an Apify Schedule. Keep the monitoring scope unchanged between runs.

How do I report a problem?

Open the Actor's Issues or Support tab and include the run ID, sanitized input, expected result, and the target coverage status. Do not post API tokens or private credentials.