Scraping cost calculator
Estimate the monthly cost of extracting web content across the major scraping APIs and self-hosting. Slide your volume, pick whether pages need JavaScript rendering and what proxy tier you need, and read the honest comparison below.
Cheapest Keirolabs at $26.32/mo for 100,000 pages. Content-extraction endpoint. Render + residential proxy bundled into one flat per-page price.
| Provider | Per page | API cost/mo | Infra/mo | Total/mo | Notes |
|---|---|---|---|---|---|
| Keirolabs | $0.0003 | $26.32 | — | $26.32 | Content-extraction endpoint. Render + residential proxy bundled into one flat per-page price. |
| Firecrawl | $0.0032 | $336.84 | — | $336.84 | 1 credit per page. Render + proxy bundled; cost is constant regardless of render or proxy. |
| Self-hosted (Playwright) | $0.0040 | $421.05 | $20.00 | $441.05 | Your VPS (~$20/mo) + proxy bandwidth + your own dev/maintenance time (not priced here). No per-page API fee. |
| Apify | $0.0060 | $631.58 | — | $631.58 | Actor compute units + proxy add-ons. Residential proxy is a priced add-on. |
| ScraperAPI | $0.0085 | $894.74 | — | $894.74 | Credit cost scales up for JS render and premium/residential proxies. |
| ScrapingBee | $0.080 | $8,421 | — | $8,421 | API credits consumed scale with JS render and proxy tier. Residential is premium. |
How the model works
Each provider charges per successful page. If you expect a 95% success rate, you will actually fire 105,263 requests to net 100,000 pages — so the model divides your target volume by the success rate before multiplying by the per-page price. Self-hosted adds a fixed ~$20/mo VPS line and charges you for proxy bandwidth but nothing per page to an API vendor.
Per-page figures are transparent estimates taken from each provider’s published pricing as of August 2026. Keirolabs bundles JavaScript rendering and residential proxies into one flat per-page extraction price, which is why it stays constant across the render and proxy toggles above. Managed scraping APIs (ScraperAPI, ScrapingBee, Apify) charge extra credits for JS render and for residential proxy tier — toggle those above to watch the cost move. Verify current prices on each provider’s site before you commit budget.
When each option actually wins
- Keirolabs wins when your goal is clean page content for RAG or search — flat $0.25/1k regardless of render or proxy, no infra to run. It is a content extraction API, not a full crawl-and-control suite.
- Firecrawl wins when you need a full crawl + extract + structured-data pipeline from one API. One credit per page, predictable.
- ScraperAPI / ScrapingBee / Apify win when you need fine-grained control over proxy tier, headers, and render — and accept that those knobs raise the per-page price.
- Self-hosted wins on raw per-page API cost (zero) but only if you already run infra, already own proxy capacity, and can absorb the engineering time to keep a headless fleet alive. At scale on anti-bot targets, the hidden cost is dev time, not dollars.
Further reading
- Scraping at scale: queues, concurrency, and cost
- Web scraping APIs for AI agents
- Residential proxies for scraping
- Bypassing anti-bot protections
Prices are estimates for planning, not quotes. Success-rate retries, cache hits, and provider credit bundles will change your real bill. Built by webscraping.space.