G2

Jacob W.

Owner

Small-Business (50 or fewer emp.)

9/16/2026

Business partner of the seller or seller's competitor, not included in G2 scores.

"Ended our self-hosted scraper headaches for good"

4.5/5

What do you like best about Scrapfly?

The anti-bot bypass just works. Our target site runs Vercel's BotID/Kasada challenge, and a single asp=true&render_js=true request clears it reliably from any IP, no residential proxy reputation to manage and no browser-fingerprint tricks to maintain ourselves. That let us retire a self-hosted Mac that had become a single point of failure for our whole scraping pipeline and move everything onto plain GitHub-hosted runners.

Billing is also easy to reason about: it's a flat per-request credit cost with a hard monthly cap rather than a metered bill, so we can budget it like a fixed subscription instead of worrying about a surprise invoice. Small things like wait_for_selector also gave us real control over rendering timing that we didn't have when we were driving Playwright ourselves. Review collected by and hosted on G2.com.

What do you dislike about Scrapfly?

Billing for images is disproportionately expensive compared to plain page fetches. A rendered HTML page fetch costs about 30 credits, but fetching an image on the same domain runs roughly 225 credits, because Scrapfly bills every resource independently instead of capturing a whole rendered page in one pass the way a headless browser does. That forced us to model costs carefully and stagger our fetch schedule instead of just scraping everything as often as we'd like.

We've also hit real intermittent reliability issues on individual requests: occasional malformed-redirect responses where the returned URL's hostname doesn't match the path we requested, a "success" response that comes back with an empty body, and read-timeouts on image fetches even at a generous 60-second limit. None of these happen constantly, but they happen often enough that we had to build our own retry logic around them rather than trust a single request to complete cleanly. In at least one case, a failure response was also structured in a way that made a real error look like an unrelated decoding bug on our side, so we had to change our error handling just to see the actual cause.

Documentation around the free tier's credit allowance was also inconsistent. What we read as "credits never expire" didn't match what the account dashboard actually showed once we used them all up, which described a monthly renewal instead. We never got a fully clear answer on which was correct. Review collected by and hosted on G2.com.

What problems is Scrapfly solving and how is that benefiting you?

Before Scrapfly, our scraping pipeline depended on a self-hosted machine running headless Chromium to get past a site's anti-bot protection (Vercel's BotID/Kasada challenge). That created two compounding problems. First, the anti-bot layer itself was an ongoing arms race: a fresh, cookie-less browser session on every run eventually got that machine's IP reputation flagged and blocked outright, which we only fixed with persisted sessions, randomized startup delays and referer headers, none of which felt like a stable long-term fix. Second, every scraper we ran was competing for time on that one physical machine, so scheduled jobs that were supposed to run every 5 minutes were sometimes landing every 4-5 hours instead, and the whole pipeline had a single point of failure sitting in one person's home.

Scrapfly solves both at once. Its anti-bot handling clears the same challenge reliably without us managing IP reputation or browser fingerprints ourselves, and because it's just an API call, our scrapers moved onto ordinary elastic GitHub-hosted runners instead of one physical machine, which eliminated the contention and the single-point-of-failure risk entirely. The direct benefit is that a pipeline feeding real-time news and results into a live app with push notifications now runs unattended and predictably, instead of quietly degrading for a week before anyone notices a stale data feed. It also freed up engineering time that used to go into babysitting browser-automation workarounds, so that effort goes into the product instead of into fighting bot detection. Review collected by and hosted on G2.com.