Every scraper works until the source changes. What separates a tool you can build on is what happens after it breaks.

When a source blocks us or returns nothing, you don't get a cryptic failed run. You get a labeled diagnostic explaining exactly what happened — and you are not charged for it. No silent failures, no paying for empty datasets.

Anti-bot-first — Playwright stealth, reverse-engineered internal APIs, session rotation and bounded retries, tuned per platform. Where a source is genuinely unbeatable without breaking its terms, we say so in the README instead of selling you a run that will fail.

You control the proxy and the cost — bring your own residential proxy and it is billed at cost, never marked up. Datacenter by default so you don't pay for bandwidth you don't need. Official-API modes (X API v2, Rakuten, Awin) when reliability matters more than coverage.

Charged on delivery — pay-per-event on results actually extracted. Partial and diagnostic outputs are labeled and excluded from billing.

25+ platforms across 10+ countries.

Global — Amazon (BR/US/EU), eBay, LinkedIn, TikTok, Instagram, YouTube, Reddit, Telegram, X, Google Maps & Trends, TradingView, Yahoo Finance, GitHub, arXiv, RSS.

Deepest Brazilian coverage on Apify — CNPJ enrichment, Mercado Livre, Reclame Aqui, iFood, OLX/QuintoAndar, PNCP public spending, and BR job boards. LATAM forks across MX, AR, CL, CO, PE, UY.

Competitive and ad intelligence · price and review monitoring · lead generation · job and ATS data · trend detection · financial signals · affiliate networks.

README, input schema and support in English and Portuguese.

Something not covered, or covered badly? Open an issue on any actor — that is how most of these were built.