All posts
Engineering · 5 min read

How tiered routing between Web Unlocker and Scraping Browser cuts cost without cutting reliability

The naive approach is expensive

The lazy default for 'crawl anything reliably' is: always launch a full browser, always render JavaScript, always screenshot. It works, but a full browser session costs meaningfully more than a standard fetch.

What actually needs a browser

A real browser session is worth its cost when a page requires JS execution to reveal content, interaction (clicking, scrolling, login), or when defenses specifically fingerprint headless clients. Everything else is often a plain HTTP fetch away from working fine.

The routing rule

DataForGAIO defaults every request to a standard unlocker-class fetch and only uses browser rendering when you ask for it. That single default is most of the unit-economics story — and it shows up on the rate card as $0.00012 vs $0.01.

See the receipt for yourself.

Start free, fetch a page, and inspect exactly what was billed and why.