Pricing pages for web scraping tools tend to bury the one thing you actually want to know: what will this cost me for the work I am going to do? Crawlbase answers that with a model that is easy to reason about once you see its three moving parts: free requests to start, payment only for results that arrive, and a cost per request that scales with how hard a site is to crawl.
This guide explains how the Crawlbase pricing model works from first principles: the free tier, why you are billed for successful requests rather than attempts, how normal and JavaScript requests consume credits differently, what the domain complexity levels mean, and how the same model carries across the Crawling API, Smart AI Proxy, and Cloud Storage. It deliberately avoids quoting dollar figures, because those live in the live calculator and pricing page where they stay current; the goal here is to give you a durable mental model you can use to estimate any plan.
Start free, no credit card
Every Crawlbase account begins with 1,000 free requests and no credit card required. That is enough to wire the Crawling API into a project, run a few hundred real pages through it, and confirm that the data comes back the way you expect before any money changes hands. You only move onto a paid plan once you have proven the tool against your own targets, which is the order most teams want: validate first, commit later.
When you do upgrade, nothing about the model locks you in. Subscriptions are commitment-free, so you can change plans or stop entirely whenever your volume changes. There is no annual contract you have to grow into and no penalty for scaling back down after a busy month.
You pay only for successful requests
The core of the model is simple: a successful request is one delivered page, and you are only charged for those. A success can be plain HTML or a fully JavaScript-rendered page, but in both cases the defining trait is that the content came back. If a request fails or the target site blocks it before a page is delivered, you are not charged for it.
This matters more than it first appears. Scraping at any real scale means dealing with sites that fight back, and on a model that bills per attempt, every block, timeout, and retry quietly inflates your bill while handing you nothing. Pay-per-success flips the incentive. The infrastructure that retries, rotates proxies, and works around blocks is on the provider's side of the line, and you only pay when that effort actually produces a page. Your cost tracks the data you receive instead of the difficulty of getting it.
It also makes budgeting honest. If you need 50,000 pages this month, you can reason about cost in terms of 50,000 successful deliveries rather than guessing at a failure rate and padding your estimate. The number you plan around is the number of pages you want, not the number of tries it takes to get them.
How credits are consumed: normal vs JavaScript requests
Not every successful request costs the same number of credits. The first lever is how the page has to be fetched. A normal request retrieves the raw HTML a server sends, which is enough for sites that deliver their content directly in the initial response. A JavaScript request runs the page in a real browser environment so that content rendered by client-side scripts (the kind you would otherwise see only after the page finishes loading) is present in what comes back.
Rendering a page in a browser is more expensive than fetching static HTML, so JavaScript requests consume more credits than normal ones. The practical guidance follows naturally: use normal requests where the data is already in the HTML, and reach for JavaScript rendering only on sites that genuinely need it. If you are not sure which a target requires, our guide on how to crawl JavaScript websites walks through how to tell the difference and when rendering is worth the extra cost.
Domain complexity levels
The second lever on cost is the target site itself. Websites are not equally hard to crawl: some serve a stable page to anyone who asks, while others run layered anti-bot systems that actively detect and block automated traffic. Reliably getting a page back from the latter takes far more infrastructure, so Crawlbase groups domains into complexity levels and prices requests to reflect the resources each level demands. Estimating your costs starts with knowing where your targets fall.
Standard
Standard domains have stable structures and minimal or no anti-bot protection. They are straightforward to crawl, need no special handling, and are served efficiently with standard infrastructure, which makes them the most cost-effective category to request.
Moderate
Moderate domains use lighter anti-bot measures or depend on dynamic, JavaScript-rendered content. Crawling them reliably calls for extra resource layers and smarter request handling to keep success rates steady, so a request against a Moderate domain costs more than one against a Standard site.
Complex
Complex domains actively detect and block automated traffic using techniques like browser fingerprinting, rate limiting, and CAPTCHAs. Getting a page back from them depends on adaptive proxy rotation, AI-driven request strategies, and considerably more infrastructure than Standard or Moderate sites require. The most heavily defended domains, the ones that invest the most in blocking automated access, sit at the top of this range and need continuously trained models, premium proxy networks, and active monitoring to stay reliable.
Fixed
Fixed is a special category for domains that require a dedicated crawling setup with fixed pricing. LinkedIn is the headline example. Requests here work differently from standard crawling: they use trained AI bots that mimic human behavior, rely only on premium residential proxies, and often need multiple retries per request because success rates are inherently low. The fixed-price structure exists precisely because these domains do not behave like the tiered categories above.
Knowing a target's complexity level is the most reliable way to estimate your crawling costs before you commit. The pricing page and the on-site pricing calculator let you combine request type and complexity into a concrete monthly figure for your own usage.
The Crawling API is where this model lives. You send one request naming the page you want, and Crawlbase handles JavaScript rendering, proxy rotation, and CAPTCHAs behind the scenes, then returns the page only when it succeeds. You pay for delivered pages, not failed attempts, and your first 1,000 requests are free with no card.
Monthly or yearly billing
Plans can be billed monthly or yearly, and the yearly option is discounted relative to paying month to month. If your scraping is a steady, ongoing need, the annual plan lowers your effective per-request cost; if your usage is seasonal or still being figured out, monthly billing keeps you flexible. Because subscriptions are commitment-free either way, the choice is about which billing cadence fits your cash flow, not about getting locked in.
For the current numbers behind both options, the pricing page shows live tier amounts and the yearly discount, and the calculator turns your expected request mix into a monthly estimate. Those figures are kept up to date in one place on purpose, which is why this article points you there rather than quoting amounts that would go stale.
One model across the products
The same successful-request, complexity-aware logic carries across the Crawlbase product line, so understanding it once tells you how to read pricing for each tool.
Crawling API
The Crawling API is the direct expression of the model. Domain complexity levels apply to it directly, JavaScript requests cost more than normal ones, and you are billed only for delivered pages. It is the right starting point for most projects that need pages back as HTML.
Smart AI Proxy and AI Proxy
The Smart AI Proxy exposes the same crawling capability through a standard proxy endpoint, so you can route existing tools through Crawlbase without rewriting them to call an API. Because it runs on the same crawling infrastructure, its pricing reflects the same complexity-driven costs underneath. If you are weighing a proxy-style integration against the API, our breakdown of why teams choose Crawlbase over raw proxies and competitors covers the trade-offs.
Cloud Storage
Cloud Storage holds the pages you crawl so you can fetch them again without re-crawling. It has its own free tier of up to 10,000 stored documents with 14-day retention, which is generous enough to cover plenty of projects outright. Beyond that, storage scales on its own plan, separate from your crawling spend.
Estimating your cost and going bigger
Putting the pieces together, a Crawlbase estimate comes down to three questions: how many pages do you need, do those pages require JavaScript rendering, and how defended are the target domains. Multiply expected successful requests by the per-request cost for your request type and complexity level, and you have a monthly figure. The on-site calculator does exactly this arithmetic for you, and the pricing page carries the live tier amounts that feed it.
If your volume runs past the published tiers, or you need a dedicated arrangement for a high-complexity target, you can contact the team for higher-volume and enterprise options. For broader context on choosing a managed scraping service over rolling your own, see our guides on the best web scraper API in 2025 and how to scrape websites without getting blocked.
Key takeaways
- Start free, stay flexible. Every account gets 1,000 free requests with no credit card, and paid subscriptions are commitment-free, so you can stop or change plans anytime.
- You pay for successes, not attempts. One successful request equals one delivered page; failed or blocked requests are not charged, so cost tracks the data you actually receive.
- Request type changes the cost. Normal HTML requests consume fewer credits than JavaScript-rendered requests, so render only when a site genuinely needs it.
- Domain complexity sets the rate. Standard, Moderate, and Complex domains cost progressively more, and Fixed domains like LinkedIn use a dedicated setup with fixed pricing.
- One model, every product. The same logic spans the Crawling API, Smart AI Proxy, and Cloud Storage; check the pricing page and calculator for live numbers and contact the team for higher volume.
Frequently Asked Questions (FAQs)
Is Crawlbase free to try?
Yes. Every account starts with 1,000 free requests and no credit card is required. That is enough to integrate the Crawling API and run real pages through it so you can confirm it works against your own targets before paying for anything.
What counts as a successful request?
A successful request is one delivered page, whether it is plain HTML or a fully JavaScript-rendered page. The defining trait is that the content came back. Requests that fail or are blocked before a page is delivered are not charged, so you only pay for pages you actually receive.
Why do JavaScript requests cost more?
A normal request fetches the raw HTML a server returns, while a JavaScript request runs the page in a real browser environment so script-rendered content is present in the response. Running a browser is more resource-intensive than fetching static HTML, so JavaScript requests consume more credits. Use them only on sites that genuinely need rendering.
What are domain complexity levels?
They are categories that reflect how hard a site is to crawl. Standard domains have little or no anti-bot protection, Moderate domains use lighter defenses or dynamic content, and Complex domains run layered systems like fingerprinting and CAPTCHAs that take much more infrastructure to get past. Fixed is a special category for domains such as LinkedIn that need a dedicated setup with fixed pricing. A request costs more on a more defended domain.
How much does a plan cost?
Tier amounts change over time and depend on your request type and the complexity of your targets, so the current figures live on the pricing page and in the on-site calculator rather than in this article. Enter your expected request mix into the calculator and it returns a monthly estimate; the pricing page shows live tier amounts and the yearly discount.
Can I cancel anytime, and what about high volume?
Yes. Subscriptions are commitment-free, so you can change plans or stop whenever your needs change, with no annual contract required. If your volume runs past the published tiers or you need a dedicated arrangement for a high-complexity target, you can contact the team for higher-volume and enterprise options.
Crawl any site at scale, without fighting infrastructure.
Crawlbase handles proxies, fingerprints, and CAPTCHAs so your team ships data pipelines instead of maintaining crawl plumbing. 1,000 requests free, no card required.