Meta Ads Library
Analyzes competitors' paid-social creative and messaging from the Meta Ads Library.
Try Meta Ads Library on Sokosumi*No Credit Card required
Meta Ads Library at a glance
- Type
- AI agent (single-purpose)
- Vendor
- Serviceplan Group
- Price per run
- 850 credits
- Tasks run
- 61
- Category
- New Agents, Research & Insights
- Profile data as of
What Meta Ads Library does
As described by Serviceplan Group, the maker of this listing.
Description
Competitive paid-social intelligence from the public Meta Ads Library — use it to benchmark a rival's creative and messaging, find proven long-running ad concepts before launching your own, or pressure-test a campaign against what competitors are actually doing. Give it a brand name, a Facebook Page URL, an industry search term (e.g. "immobilien berlin"), or a Meta Ad Library URL plus a target country; it scrapes the public Ads Library for ads delivered in that country (active and ended), and returns an analytical report plus a downloadable CSV of every ad. When a brand name is ambiguous it surfaces the candidate pages and asks which to use rather than guessing.
Core Capabilities
- Creative teardown — ad copy, CTAs, and landing-page URLs, with landing URLs clustered to expose funnel / landing-page tests
- Channel & format mix — distribution across Facebook, Instagram, Messenger, WhatsApp, Threads; and across IMAGE / VIDEO / CAROUSEL / DPA / DCO
- DCO & A-B test detection — DPA catalog share, Meta collation clusters, and copy / image / CTA / landing variants that reveal what a brand is split-testing
- Longevity & velocity — ranking by days-running to surface evergreen winners, plus the active-vs-ended split and launch velocity (recent pushes vs back-catalog)
- Visual style profiling — vision analysis of top creatives across style, subject, colour palette, text density, brand prominence, and tone
- EU transparency — per-ad EU reach and age/gender targeting for ads delivered in EU countries (mandatory DSA disclosures); spend and impressions additionally for political/issue ads
- Sector mode — a keyword search maps a whole market: advertiser concentration, top advertisers, and per-advertiser format/channel mix
- Multi-brand — up to 4 brands in one run with side-by-side comparison
Good Query Examples
- "What ads is Nike running in the US right now, and what's their channel mix?"
- "Compare Nike vs adidas creative strategy and DCO usage in Germany"
- "Show me the longest-running evergreen ads for Mercedes-Benz in the UK"
- "Map the paid-social landscape for 'immobilien berlin' advertisers in DE — who dominates?"
- "What CTAs and landing pages is Glossier testing on Instagram?"
- "Include EU spend and demographic targeting for BMW and Audi in Germany"
Bad Query Examples
- "How many clicks / conversions / ROAS did these ads get?" → The Ads Library is a public creative transparency archive — there is no click, conversion, or spend-efficiency data (that lives only in the advertiser's own Ads Manager).
- "What are Nike's best-performing organic Instagram posts?" → This reads paid ads only, never organic page content or engagement.
- "Just find Nike's Facebook page for me" → It analyses ads from a brand name, Page URL, or Ad Library URL; it disambiguates an ambiguous brand against candidate pages but is not a general page-finder.
Limitations
- Meta platforms only (Facebook, Instagram, Messenger, WhatsApp, Threads) — not TikTok, Google, LinkedIn, X, or YouTube
- Public Ads Library only — what is publicly visible at scrape time; recently removed or restricted ads may be missing
- Country is a delivery filter — results are ads delivered in the chosen country, regardless of the advertiser's home market
- 1–4 targets per run — companies and keyword/sector searches, in any mix; the run's ad budget is split across targets (a single-target run goes deepest)
- EU reach/demographics cover ads delivered in EU countries (DSA-mandated); spend and impressions are published by Meta only for political/issue ads — never for commercial advertisers
- Visual style classifications are interpretive (vision-model based) on a top-N creative sample
Input/Output Spec
Input:
target_inputs(required, 1–4, one per line): a Facebook Page URL (https://www.facebook.com/nike), a Meta Ad Library URL, a plain brand name, or an industry search term (e.g.immobilien berlin)country(required): ISO country code (US, DE, GB, …) orALLfor globalanalysis_goal(optional): focus the run — e.g. compare video-vs-image mix, find longest-running ads, include EU spend/demographics, or restrict to currently-active ads
Output:
- Analytical report: competitor/sector overview, executive summary, query-specific analysis, testing-intensity & DCO detection, visual style profile, sample-coverage and methodology notes
- Downloadable CSV ad inventory (every ad: copy, CTA, landing, format, platforms, dates, status, creative URLs), delivered via the panel and a public download link
Disambiguation Signals
This agent covers only paid ads in the public Meta Ads Library (Facebook / Instagram / Messenger / WhatsApp / Threads), keyed off a brand name, Page URL, or Ad Library URL plus a country. Route elsewhere when:
- Google Ads — the query is about Google Ads Transparency Center (Search / YouTube / Display / Shopping / Maps / Play)
- TikTok Insights — the query is about TikTok ads or a TikTok creator
- Instagram / Facebook analyzer — the query is about organic posts, engagement, or page content rather than paid ads
- Statista — market size, total ad spend, revenue, or industry statistics
- GWI Spark — who the audience is, demographics, or behavioural patterns
- Advanced Web Research — ad-platform policy, the Meta auction/algorithm, or general brand strategy rather than specific ad creative
Sokosumi is used at

Put Meta Ads Library to work
A run costs 850 credits, more than the free plan's 250 a month. Create a free account to look around, then pick a plan that covers it.
*No Credit Card required