How to export Meta Ad Library data

The Meta Ad Library has no export button, and most of what ranks for this query either sells you a scraper or quietly answers a different question (exporting your own Ads Manager data). This is the straight version: the four genuine routes to getting Ad Library data out of Meta, exactly what the official API returns and withholds, where scrapers stand legally and practically, and, drawing on the 14,757 Meta creatives we've AI-qualified ourselves, why a raw dump still isn't the thing you actually want.

The short answer

The Ad Library interface is view-only. You can search, filter and page through any advertiser's live ads, but there is no download, no CSV, no "save results" anywhere in the UI. If you want the data out, there are exactly four routes:

Each is covered properly below, because the differences are where every other article on this topic goes wrong.

First, the answer most posts give you is for a different question

A large share of what ranks for "export Meta ad library data" actually explains how to export reports from Ads Manager — your own account's performance data. That's a real feature (Ads Manager has an Export button; the Ad Library does not), but it only works for ads you ran yourself. If you're researching a competitor, Ads Manager is irrelevant: you have no access to their account and never will. Everything below is about the public Ad Library — other people's ads.

Route 1: the Ad Library API — official, free, and narrower than it sounds

What it actually covers

The API's single endpoint, ads_archive, searches the same archive as the website. But its scope rule is the detail almost nobody quotes, so here it is from Meta's own reference documentation:

"Ads that did not reach any location in the EU will only return if they are about social issues, elections or politics."

Read that twice if you're a UK marketer. The UK is not in the EU, so a commercial campaign targeted only at the UK is not retrievable through the API at all — it's visible in the Ad Library website, but the official export route won't return it. You get a brand's ads via the API only where the campaign also delivered somewhere in the EU (common for big multi-market advertisers, rare for UK-only ones) or where the ad is political. The endpoint also supports category filters (ALL, EMPLOYMENT_ADS, FINANCIAL_PRODUCTS_AND_SERVICES_ADS, HOUSING_ADS and POLITICAL_AND_ISSUE_ADS), but the EU-reach rule applies regardless.

What it returns

For ads in scope, you get genuinely useful structured fields: the ad body text, link titles and descriptions, start and stop dates, which platforms the ad ran on, and the advertiser's Page. The EU's Digital Services Act forced extra disclosure onto EU-delivered ads, so those also carry transparency fields such as eu_total_reach (aggregate EU audience over the ad's run), beneficiary_payers (who paid and who benefits), targeted age, gender and location parameters, and a reach breakdown by age, country and gender. For political and issue ads you additionally get spend and impressions — but only as bands (the same ranges you see in the public Ad Library Report), never exact figures.

What it withholds

• The creative itself. You get an ad_snapshot_url — a link that renders the ad in a browser — not downloadable image or video files.
• Any performance data for commercial ads. No spend, no impressions, no clicks, no engagement. Banded spend exists for political ads only.
• UK-only commercial campaigns, per the scope rule above.
• Anything analytical. No notion of offer, discount, hook or theme — it's an archive, not an analysis.

Access and rate limits

Access requires a Meta for Developers account, an app, a user access token with the ads_read permission, and, because the archive includes political ads, identity confirmation with government ID — the same process required to run political ads. Budget a few days for that. Rate limits are dynamic and unpublished: Meta's documentation specifies only the error you'll hit (613: Calls to this api have exceeded the rate limit), and in practice sustained pulls get throttled quickly, so a full export of a large advertiser is a patient, paginated job rather than one request.

Route 2: the Ad Library Report — the only official download button

Meta does publish downloadable files: the Ad Library Report, which lists advertisers and their banded spend by country and date range. The limitation is total: it covers only ads about social issues, elections or politics. For journalists and researchers tracking political spend it's genuinely good. For commercial competitor research it contains nothing — your rival's Black Friday campaign will never appear in it.

Route 3: Meta Content Library — the researcher route

The Meta Content Library and its API give much deeper programmatic access to public Facebook and Instagram content, free of charge — but only to researchers affiliated with academic institutions or not-for-profits doing scientific or public-interest research, working inside a controlled environment. If that's you, it's the best-kept route on this list. If you're a marketer, you won't qualify, and the access terms prohibit commercial use anyway.

Route 4: scrapers — what the listings don't tell you

This is what most of the first page for this query is selling: hosted scrapers that paginate the Ad Library website and emit JSON or CSV per ad. Three things the listings underplay:

• It's against Meta's terms of service. Automated collection without permission breaches Meta's terms. The practical risk mostly lands on the scraper operator, but you're building a workflow on sand.
• They break, often. The Ad Library's markup changes regularly and scrapers lag behind each change. We run our own collection pipeline across four ad platforms and maintenance is a permanent line item, not a one-off setup cost; nobody selling a $10 scraper mentions this.
• The output is still raw. You get what's on the page: text, dates, a media link. No matter which scraper you pick, you've exported the same unanalysed material the API gives you, minus the legitimacy.

Why a raw export still isn't analysable

Here's the part we can speak to directly, because we've done the next step at scale. Suppose you solve export perfectly: every live ad for your five competitors lands in a spreadsheet — ad text, dates, snapshot links. Now try to answer the questions you actually had:

• Are they discounting right now, and how deep?
• What hooks are they leading with?
• Which ads are long-running bankers versus fresh tests?
• Who are they talking to?

None of that is a column in any export, official or scraped. It lives inside each creative — in the image, the video, the copy — and has to be read out ad by ad. That reading step is what we do: AdRival captures ads for 164 UK brands across Meta, Google, TikTok and LinkedIn and has AI-qualified more than 50,000 creatives, scoring each one on 19 fields — offer type, discount percentage, hook, target audience, urgency and more. On Meta specifically that's 14,757 qualified creatives from 146 brands, and the qualified layer shows things no raw dump can:

• 4,350 of those 14,757 Meta creatives (29%) carry an explicit offer — meaning 71% of what you'd export is brand, launch or lead-gen work with no offer to find.
• Among offer-led ads, percentage discounts dominate (1,471 ads), followed by free trials (898), money-off amounts (700), free gifts (476), bundles (321), finance offers (314) and free delivery (148).
• 1,978 ads state a specific discount percentage, 538 name an offer deadline, and 192 expose a promo code — fields that simply don't exist until someone reads the creative.

If you want to see what qualified output looks like rather than take our word for it, our brand pages are generated live from this pipeline: compare the offer-heavy mix on Butternut Box's live ads with the hook-led approach on Trinny London's, or the sheer volume a brand like Vets4Pets keeps live. Each page is a free teaser of the structured export this guide is about.

A sensible workflow for UK marketers

1. Occasional spot checks: just use the Ad Library website. It costs nothing and the view-only limitation doesn't matter for a quick look. Our Meta Ad Library guide covers how to search it well, and why ads sometimes don't show.
2. Political or EU-wide research: the official API and the Ad Library Report are legitimate and free. Verify your identity early — it's the slow step.
3. UK commercial competitor research: the API won't return UK-only campaigns and scrapers hand you unqualified raw material. Either accept manual reading, or use a service that collects and qualifies — that's the product we sell, and the honest framing is that the qualification, not the export, is the hard part. Our Meta Ad Library alternatives guide compares the paid options, including where ours isn't the right fit.

One caveat on all exported Ad Library data, including ours: the library shows what's live, not what's working. An exported ad with no performance signal is a hypothesis, not a result — treat long run-times and repeated re-launches as the closest available proxy for success.

Related guides

• Meta Ad Library: how to use it — and what it actually shows
• Meta Ad Library alternatives: the honest UK guide
• Instagram Ad Library · TikTok Ad Library · Google Ads Transparency Center