scrapeprices 
Crawlbase
The cheapest way in of any general scraper here and the hardest one to budget. Entry is $0 with no minimum, the pay-as-you-go card starts at $3.00 per 1,000 and falls to $0.02 per 1,000 at the top bracket, subscriptions start at $99 for 200,000 credits, and only successful requests are billed in Crawlbase's own words. Residential exit nodes, CAPTCHA solving and Cloudflare, PerimeterX and DataDome handling carry no surcharge at all, which is unusual in this market where those are the two big multipliers. The catch is the multiplier you cannot see: every domain is silently assigned a difficulty tier from Standard x1 up to Extreme II x20, and Crawlbase never publishes which tier any site is in, so the anti-bot price is unknowable before you send the request - you read it back afterwards from the domain_complexity field or the X-Domain-Complexity header. JavaScript rendering is a flat x2 on top of whatever that tier turns out to be. Two disagreements worth knowing: the pricing page offers "up to 5,000 requests free" in one block and "1,000 free credits" in another, and it advertises "UPTIME 99.99%" without publishing an SLA that would make the number a commitment.
Plain page, by volume
| Pages per month | Monthly bill | Per 1,000 | Cheapest route | Allowance unused |
|---|---|---|---|---|
| 1k | $2.00 | $2.00 | pay as you go | - |
| 10k | $20.00 | $2.00 | pay as you go | - |
| 100k | $60.00 | $0.600 | pay as you go | - |
| 1M | $349 | $0.349 | Scale 1M | - |
How Crawlbase compares
At 10,000 pages a month Crawlbase costs $2.00 per 1,000, 15.4x the cheapest published route and 2.9x AnyAPI.
| Provider | 1k a month | 10k a month | 100k a month | 1M a month |
|---|---|---|---|---|
Z  Zyte API | $0.130 | $0.130 | $0.130 | $0.100 |
A  AnyAPI | $0.700 | $0.700 | $0.700 | $0.700 |
S  scrapestack | $19.99 | $2.00 | $0.200 | $0.080 |
| C Crawlbase | $2.00 | $2.00 | $0.600 | $0.349 |
O  Olostep | $3.33 | $2.00 | $0.990 | $0.399 |
List prices, dollars per 1,000 Plain pages at the cheapest published route to each volume. How this is worked out
Every published plan
| Plan | Price | Included credits | Overage each | Note |
|---|---|---|---|---|
| Pay as you go | $0 per month | 0 | $0.0030 | No monthly fee and no minimum. Graduated ladder; the rate here is the first bracket, $3.00 per 1,000 standard credits. See payAsYouGoTiers for the whole card. It is a rate per credit, so the JavaScript and site-difficulty multipliers apply on top. |
| Starter | $99.00 per month | 200,000 | not published | Yearly billing is one month free plus 8% off. Overage continues on the graduated pay-as-you-go card at the bracket already reached, so there is no single published overage rate. |
| Scale 500K | $199 per month | 500,000 | not published | Overage continues on the graduated card. |
| Scale 1M | $349 per month | 1,000,000 | not published | Marked "Most popular". Overage continues on the graduated card. |
| Growth | $599 per month | 2,000,000 | not published | Overage continues on the graduated card. |
| Enterprise | on request | - | not published | Contact-us tier: custom volume and invoicing, no public price. |
"One credit is one standard request." JavaScript rendering is a published x2. Residential is 1 because Crawlbase does not sell it as a tier at all: "Rotating residential proxies, CAPTCHA and block handling are part of every request, never a separate bill", and the Crawling API docs say "The request is sent through a residential or datacenter exit node - automatically by default", with no parameter to force one and no surcharge. Markdown is 1 for the same reason: the docs expose format=md ("format=md converts the page to GitHub-Flavored Markdown", pair with md_readability=true) and it appears nowhere in the published multiplier list, which otherwise itemises JavaScript x2, screenshots at 2 credits and cloud storage at 0.5. Anti-bot is null because its price is genuinely unpublished: challenges are solved server-side ("Cloudflare, PerimeterX, DataDome, hCaptcha, and other common challenges are solved server-side") but the bill is set by an automatically applied site-difficulty tier - Standard x1, Moderate x1.5, Complex x2, Complex I x3, Complex II x4.5, Extreme x9, Extreme I x13.5, Extreme II x20 - and Crawlbase never says which tier any named site falls into. You learn the tier after the fact, from the domain_complexity field and the X-Domain-Complexity header. Scroll requests are billed by time: "first 8 seconds = 1 request; each additional 5 seconds adds 1 more".
Cost by kind of fetch
Per 1,000, at 10,000 pages a month.
What it can do
Only what the provider's own pages state.
Where these numbers came from
- https://crawlbase.com/pricing read 2026-09-14 / Plans $99/200K, $199/500K, $349/1M, $599/2M plus Enterprise contact-us; pay-as-you-go ladder $3.00, $2.00, $0.60, $0.50, $0.10, $0.05, $0.04, $0.02 per 1,000; "One credit is one standard request"; JavaScript x2, screenshot 2 credits, cloud storage 0.5 credits; difficulty ladder Standard x1 to Extreme II x20; yearly is one month free plus 8%. Verbatim: "Only successful requests are billed. A request that fails or is blocked costs nothing, whatever the site or rendering mode." Verbatim: "Rotating residential proxies, CAPTCHA and block handling are part of every request, never a separate bill." Verbatim on expiry: "Pay as you go bills in arrears for what you actually used, so there is nothing to expire. A subscription's credit block resets with each monthly cycle." Also the two conflicting free grants, "Up to 5,000 requests free. No credit card." and "Every new account starts with 1,000 free credits and no card on file", and a "UPTIME 99.99%" figure with no SLA document behind it. Markdown is not mentioned on this page at all.
- https://crawlbase.com/docs/crawling-api/parameters/ read 2026-09-14 / format=md ("converts the page to GitHub-Flavored Markdown") with md_readability=true, X-Markdown-Flavor / X-Markdown-Features / X-Markdown-Base-URL response headers; country= two-letter ISO geo routing; tor_network=true; "The request is sent through a residential or datacenter exit node - automatically by default"; "Cloudflare, PerimeterX, DataDome, hCaptcha, and other common challenges are solved server-side. You get the post-challenge HTML, not the challenge page"; billing only on cb_status 200; scroll billed as first 8 seconds = 1 request, each further 5 seconds = 1 more. No credit cost is attached to format=md anywhere on the page.
Something out of date? Send the provider's own page to hello@scrapeprices.com and it is corrected on the next capture.