Some links on this site are affiliate links: if you buy through them we may earn a commission, at no extra cost to you. It never changes our rankings or numbers. How this site earns.
Methodology
Where the prices come from
Every plan price and credit multiplier is read from the vendor's own pricing page or documentation and checked against the page it came from.
Every scraping API plan, credit multiplier and proxy price, with each vendor's pricing page and the date we read it, is also a spreadsheet you can download from the comparisons page.
From pages to a price
- Credit-based APIs: successful pages × the vendor's credits for that kind of page, then the cheapest plan whose monthly credits cover it.
- A plan's own price per 1,000 pages, in the plan tables on each API's page, assumes its monthly credits are used in full, with monthly billing.
- Comparison tables show each API's cheapest published plan for the month, billed monthly. "Custom" means the volume needs more than its largest published plan.
- Comparisons across workloads use four monthly volumes (10,000, 50,000, 250,000 and 1,000,000 successful pages) and four kinds of page: plain HTML, JavaScript rendering, premium proxies, and both.
- Bright Data Web Unlocker: one price per successful request, from pay-as-you-go or a plan plus its stated rate for extra requests.
- Oxylabs: a plan's fee is drawn down at its per-1,000 rate for general sites, with a higher rate for JavaScript rendering.
What we leave out
- Overage and top-up credits: we price whole plans, which can overstate the cost just above a plan's limit.
- Per-site surcharges (for example Google or LinkedIn on ScraperAPI and Scrape.do) and difficulty multipliers (Crawlbase, Zyte). Google results pages are priced on their own in SERP API pricing, with each search API's plans and the scraping APIs' credit cost for a Google page.
- Zyte, priced by each website's difficulty tier, and Apify, priced by compute, can't be converted to a fixed price per page. Their own pages lay out what they do publish: Zyte pricing and Apify pricing.
- Failed requests: most vendors don't charge for them, so we price successful pages only.
Free plans and trials are counted in pages at each API's own credit cost per page, from its pricing page or documentation.
Proxy prices
Proxy prices come from 13 providers' own pricing pages, read on 6 October 2026 (DataImpulse's on 16 October 2026, Evomi's on 10 October 2026 and Oxylabs' on 7 October 2026): by plain HTTP, or in a plain headless browser where a page draws its prices with JavaScript. Every figure is saved with the exact page text it came from, and each quote was checked against the saved page.
- Plans with an allowance
- So many GB or IPs a month: the cheapest plan whose allowance covers what you need, at its full price. Top-ups whose price the page doesn't state aren't used, which can overstate the cost just above a plan's size.
- Pay as you go and volume tiers
- The per-GB or per-IP rate for an order that size. Below the smallest order you pay for the smallest order, unless the provider says purchased traffic never expires; then only what you use in the month counts, and the minimum first purchase is named next to the price.
- Monthly commitments
- Bright Data, Evomi, Oxylabs and SOAX: you pay the commitment or your usage at the plan's rate, whichever is more, as their pricing pages describe.
- Per GB
- The month's cost divided by the GB you use, so a plan bigger than you need costs more per GB.
- Comparison tables
- The month's cost on each provider's cheapest published option for that much traffic or that many IPs, before VAT, with monthly billing. "Not sold" means its pricing pages don't list that type; "By quote" means the volume is above its largest published plan.
- Discounts the page applies without a code
- In the prices: where a page shows a struck-through price beside the price charged, the price charged is used. Coupon codes and first-months offers are shown beside the price but not used in rankings.
- Bandwidth calculator
- The bandwidth calculator turns pages into GB at a page size you choose, by default HTTP Archive's 2025 median page sizes, taking 1 GB as 1,000,000 KB. Headers, retries and blocked pages add traffic it doesn't count.
- Left out
- VAT, yearly billing, SOAX's cheaper rates outside its tier 1 countries (which include the United States), "starting from" prices with no volume attached, and per-IP prices that don't include traffic.
NetNut isn't priced. When we checked, netnut.io showed only a page titled "Seized by the Federal Bureau of Investigation" instead of its website, so no prices could be read. We don't fill gaps from reviews or memory.
Unlockers, browsers, search APIs, CAPTCHAs and Google Maps
- Web unlockers
- Plans sold by requests are priced like scraping APIs; plans sold by GB from your requests times the response size (by default the median page's HTML, 21 KB, from HTTP Archive), taking 1 GB as 1,000,000 KB; Oxylabs' plans are a balance spent per GB at the plan's rate. IPRoyal's request packages count as the cheapest mix that covers the month. Coupons and free tiers aren't counted.
- Hosted headless browsers
- Browser hours are sessions times minutes, priced on each plan's fee, the hours, units or credits it includes, and its rate for more; a plan whose longest session is shorter than yours is skipped. Proxy traffic is listed separately, because each vendor meters it differently.
- Web search APIs
- A search is one request in the vendor's default search type, with up to 10 results where the price depends on the count. Credits are converted at what a search uses (Tavily 1 credit for a basic search, Firecrawl 2 per 10 results), and plans that sell credits past their limit are priced with them. Jina's tokens are priced at its stated minimum of 10,000 a search, and a token package counts in the month it's bought, because its price table doesn't say how long tokens last. Free monthly credits aren't taken off the prices; they're listed beside them.
- CAPTCHA solvers
- Each service's published price per 1,000 solved CAPTCHAs times your volume; where a service publishes a range, it is ranked by the top of it.
- Google Maps
- Scrapers per place or record; SerpApi's searches and Google's Places API requests at full pages of 20 places, with Google's graduated tiers after its free monthly cap.
How this site earns
Links can be affiliate links: once their programmes accept the site, several of these APIs may pay us a share of what you spend if you sign up through them. Those links are labelled "Affiliate link". Programmes that have accepted the site so far: Scrape.do and SearchApi. Rankings are sorted by price alone, and APIs that pay us nothing, such as Scrapfly, are included on equal terms.
How the site is made
Prices are collected and checked by automated scripts. Every figure is computed by code from the cited data.
Prices checked against each platform's own pricing page on .