Scrape.do Independently tested Pricing: claimed
Bypass-focused scraping API that only charges for successful requests, with geolocation and JS rendering built in.
Vendor: Scrape.do · Official site ↗
Independently tested (2026-10-26)
| Target tier | Label | Success | Median ms | Errors |
|---|---|---|---|---|
| static | Static HTML (example.com) | 5/5 | 2820 | — |
| static | Static HTML (httpbin) | 5/5 | 1531 | — |
| js-optional | Regular site (this site, Cloudflare Pages) | 5/5 | 2417 | — |
| cloudflare | Cloudflare-protected (demo page) | 3/3 | 9867 | — |
Success = HTTP 200 AND validator substring present. Cost per 1k successful = claimed list $/1k ÷ success rate. Volume is intentionally tiny (benchmarking advertised capability, not load testing). Run origin matters: requests originate from our network; the no-key baseline reflects local conditions, commercial APIs fetch from their own infrastructure. Full methodology: How we measure.
Head-to-head: see Scrape.do vs ScraperAPI on the same targets — validator-checked success, latency and cost per 1k successful.
Real product pages, 3-day repeat (2026-10-23 → 2026-10-25)
| Real target | Success | Day medians (ms) | 3d median |
|---|---|---|---|
| 商品列表页 · static | 18/18 | 957 / 2488 / 2549 | 2488ms |
| 商品详情页 · static | 18/18 | 2106 / 1109 / 1390.5 | 1390.5ms |
| 真实电商详情页 · JS-render | 0/18 | — / — / — | — |
Declared cost of this run: $0.0060 (54 credits) — includes failed renders on the JS-heavy target. Success = HTTP 200 AND product content present. Full story: the 3-day head-to-head report.
Claimed pricing — copied from the vendor's public pricing pages as of 2026-10-22. Performance numbers above are our own measurements; pricing remains vendor-claimed.
| Model | credits (successful calls only) |
|---|---|
| Entry tier | 1,000 free successful calls; Hobby $29/mo ≈ $0.11/1k; Business $249/mo ≈ $0.07/1k |
| Notes | Failed/timed-out requests are not charged. ~20% off on annual plans. Verified 2026-10-22. |
Capabilities
vs. the measured ecosystem
For context: the cheapest measured pay-per-result Actor on the Apify Store bills about $0.0100 per 1,000 results at list price, and dedicated anti-bot Actors show 0.00%–3.26% measured 30-day failure rates (see the bypass report).
Before you commit — five questions to ask Scrape.do (or any scraping vendor)
- What exactly is a "request"? Scrape.do lists credits (successful calls only) pricing (1,000 free successful calls; Hobby $29/mo ≈ $0.11/1k; Business $249/mo ≈ $0.07/1k). Per-request, per-credit and per-GB models produce very different bills at the same volume — normalize to $/1k successful before comparing.
- Do failed requests bill? The most common hidden cost. In our 3-day benchmark of two providers, failed JS renders on the same target cost $2.38 vs $0.006 — the billing structure, not the list price, decided that gap.
- What do JS renders and premium IPs multiply by? 5–25× multipliers on rendering or residential exits are common market-wide. Get the full multiplier table in writing before committing volume.
- What's the success rate on your target? Vendor dashboards usually count HTTP 200 — a 200 that carries a block page is still a failure for your business. Pilot on your actual target with a content validator.
- What's the exit cost? Output formats, delivery options, and whether re-pulling historical data is included in the plan.
Frequently asked questions
What does Scrape.do cost?
Claimed entry tier: 1,000 free successful calls; Hobby $29/mo ≈ $0.11/1k; Business $249/mo ≈ $0.07/1k (pricing model: credits (successful calls only), checked 2026-10-22). In our benchmark: ~$0.11 claimed per 1k ÷ 100% measured success.
Is Scrape.do independently tested?
Yes — our own black-box run on 2026-10-26: 100% validator-checked success, 2642ms median, plus a 3-day real-product-page repeat (see the table above).
What should I double-check before signing up?
Three things in the contract, not the marketing: failure billing terms, the JS-render/residential multiplier table, and a paid pilot on your actual target site with a content validator — not a demo page.
How to read this data
Performance figures come from our own black-box benchmark run on 2026-10-26 (validator-checked success, tiny volume — see the note above). Pricing is vendor-claimed, checked 2026-10-22.