Skip to content
AdLibraryScraper.com
Menu
See the top picks

How We Rate Ad Library Scrapers

The criteria behind every score, and how we work out the price of 1,000 ads with EU reach

Oleg Glanz

By: |Updated:

Each tool is scored from 1 to 5 on 7 criteria. The overall score is the plain average of those numbers, with no hidden weighting.

The 7 criteria

  • Price per 1,000 ads: What 1,000 ads cost with EU reach data included, on the entry plan and at volume, plus minimum spend and whether credits expire.
  • Data depth & EU reach: Which Ad Library fields come back: creatives, copy, dates, platforms, EU total reach with age/gender/country breakdown, spend estimates, landing pages.
  • Reliability & maintenance: Who maintains the scraper, how fast breakages get fixed after Meta changes the Ad Library, and published success rates or SLAs.
  • Ease of use: Time from sign-up to first usable dataset, quality of docs, and whether non-developers can use it.
  • AI & MCP readiness: Official MCP server for Claude, ChatGPT or Cursor, OAuth sign-in, compact LLM-friendly responses, agent skills and documented credit cost per call.
  • Integrations & exports: REST API, SDKs, CSV/JSON exports, webhooks, Clay, Make, Zapier and n8n.
  • Free tier & trial: How much real data you can pull before paying, and whether a credit card is required.

How we calculate the price per 1,000 ads

Vendors price Ad Library data in different units: per ad, per request, per credit or per record. To compare them, we convert every price into the cost of 1,000 Meta ads, and we show two numbers.

  • Ad list only: the ads a search returns. Request-based APIs return a page of about 30 ads per search request, so 1,000 ads take about 34 requests.
  • With EU reach: the same ads plus Meta's EU transparency data (eu_total_reach and the age, gender and country breakdown). Most APIs only return it from an ad-details endpoint, one request per ad, so we add 1,000 requests. Tools that include reach in each result are priced per ad.
  • The entry price uses the cheapest paid plan or pack; the volume price uses the largest public tier. Prices exclude VAT. When a vendor does not publish a price or the reach field, we show that instead of estimating it.

How we rate AI and MCP readiness

We check whether each vendor runs an official MCP server, whether Claude and ChatGPT can connect with OAuth or need a key in the URL, how compact the responses are for an LLM context window, and whether the agent can see what each call costs. Setup commands in our reviews come from each vendor's own MCP documentation.

Where the data comes from

  • Prices, free tiers and features come from each vendor's own pricing and product pages. Every review links to those pages, and the "checked" date shows when they were last verified (November 2, 2026).
  • User counts, ratings and maintenance response times for Apify actors come from the public Apify Store listings.
  • "Brand searches" are monthly US Google search volumes from the DataForSEO Google Ads API. They show how widely a tool is known, not how good it is, and do not affect scores.

What we do not do

  • Vendors cannot pay to change a score or their position.
  • Affiliate commissions do not affect scores. Tools without an affiliate program are listed and scored in the same way.
  • We do not claim figures we cannot source. When a vendor does not publish a number, we say so instead of estimating it.

Updates

Ad Library scrapers change quickly because Meta changes the Ad Library. We re-check prices and features at least quarterly and whenever a reader reports a change. See the current ranking on the best Ad Library scrapers page.