Octoparse Pricing: Cost per Task and Plan

Almost every tool in this category bills you per page, per request or per credit. Octoparse does not, and that single difference is why its pricing page is hard to compare against anything else you are evaluating. There is no cost per 1,000 pages to calculate, because pages are not the unit.

Every figure below was read from the Octoparse pricing page, its plan comparison in the help centre and its own pricing explainer, in September 2026. None of it comes from a review aggregator, which matters here because the monthly and annual prices differ by 16 percent and aggregators routinely quote one as if it were the other.

Octoparse pricing plans

Start with the number the pricing page leads with. It advertises "from $69/mo", and that is the annual-billed rate. The month-to-month price is $83. Octoparse says so plainly in its own explainer, and the arithmetic confirms it: 16 percent off $83 is $69.72, and 16 percent off $299 is $251.16.

A task in Octoparse is a saved crawler configuration, not a run and not a page. Build a scraper for one site, save it, and that is one task against your cap. Run it every night for a year on a hundred thousand pages and it is still one task.

What does an Octoparse task actually cost?

Divide and the picture inverts depending on which limit you look at. Per task, Standard is $0.83 a month and Professional is $1.20, so the cheaper plan gives you better value per configuration. Per concurrent cloud run, Standard is $27.67 a month and Professional is $14.95, so the expensive plan is 46 percent better value on throughput.

That is unusual. In a credit-metered product the higher tier discounts the same unit, and the only question is whether you will use the volume. Here the two meters point in opposite directions, and the right tier depends entirely on which constraint binds first for you.

So the question to answer before you pick a plan is not how many pages you will pull. It is how many separate sites you need configured, and how long you are willing to wait for the queue to empty.

Is Octoparse cheaper than a web scraping API?

On a narrow, deep workload it is close to unbeatable, and that deserves saying plainly. Suppose you monitor six large retail catalogs, half a million pages a month between them. That is six tasks and unlimited export, so $69 a month billed annually covers it. Metered per page, the same half million pages costs real money almost anywhere, ourselves included.

Turn the workload sideways and it reverses. Suppose you pull thirty pages each from four hundred different suppliers. The volume is trivial, twelve thousand pages, but it is four hundred configurations. That exceeds the Professional cap of 250 and puts you in an Enterprise conversation for a job that any per-page API would bill at a few dollars. Worse, somebody has to build and maintain four hundred point-and-click scrapers, and they break when the sites change.

The crossover is breadth, not volume. Few sites and deep crawls favor the task meter. Many sites and shallow crawls favor a page meter, and the gap widens the more sources you add.

What add-ons does Octoparse charge for?

This is where "unlimited" needs qualifying. Export volume is genuinely unlimited on paid plans, but the things that make a hard site work are unbundled and metered.

• Residential proxies, $3 per GB. This is the one that quietly reintroduces a per-page cost. At roughly 1 MB of traffic per rendered page, $3 per GB is about $3.00 per 1,000 pages; at 2 MB it is about $6.00. Your real page weight decides it, so measure it on your targets before you assume the flat fee is the whole bill.
• CAPTCHA solving, $1 to $1.50 per 1,000 solves. Only charged when a solve is needed, so it is cheap on ordinary sites and a genuine line item on defended ones.
• Pay-per-result templates, $0.001 to $3 per 1,000 results. A three-thousand-fold range, so read the rate on the specific template rather than the category. You pay only for successful rows.
• Done-for-you services. Crawler Setup starts at $399 and Data Service at $599, both one-off rather than subscription.

The template library is the part that most often turns into a recurring cost, because the templates people reach for are the lead-list ones. That is worth a budget note beyond the scraping itself: a list of prospects does nothing until it reaches a human, and teams that automate the collection and then hand-sort the output in a spreadsheet usually find the bottleneck moved rather than disappeared. Routing the rows to an owner the moment they land is a solved problem with its own tooling, and it is cheaper to solve than the scraping was.

Does Octoparse have a free plan?

Yes, and it is more limited than the row count suggests. Free gives 10 tasks, 1 device, 1 user, 2 concurrent runs and 50,000 exported rows a month. The constraint that matters is that extraction is local only: no cloud runs, no IP rotation and no scheduling. Your machine has to be awake and running the app for anything to happen.

That makes it a real evaluation tier and not a production one. It is enough to confirm the point-and-click recorder can handle your target site, which is exactly the right way to use it before paying for anything.

How long does Octoparse keep your data?

Cloud run history is shorter than most buyers expect. The plan comparison lists 10 records on Standard, 20 on Professional and 100 on Enterprise. If your process depends on reaching back through past runs, plan to export on every run and keep the history in your own store rather than treating the platform as an archive.

Is there an Octoparse refund?

A 5-day money-back guarantee applies to paid plans. That is tighter than the 7 to 30 day windows common in this category, so the practical advice is to do your real evaluation on the free tier and treat the first paid week as a deadline rather than a cushion.

When is a no-code scraper the wrong purchase?

Three cases, and none of them is about price.

The first is many sources. Point-and-click configuration scales linearly with the number of sites, so four hundred suppliers means four hundred things to build and repair. An API with a schema handles new sources as a change of URL.

The second is output shape. Octoparse produces rows for a spreadsheet or a database, which is the right artifact for analysis. If the destination is a retrieval pipeline or a model prompt, you want clean markdown or typed JSON rather than a flat table, and converting one into the other after the fact loses the structure you needed.

The third is integration. If the scrape has to run inside an application rather than beside it, a desktop recorder and a cloud scheduler are the wrong shape regardless of what they cost.

The honest comparison

ClawEngine meters pages: $39 a month for 50,000 pages on Hobby, which is $0.78 per 1,000, $99 for 250,000 on Startup and $399 for 1,500,000 on Scale. One page of allowance per page returned, whether it yields one field or four hundred rows, with discovery included in the same meter.

Be clear about what you give up. There is no point-and-click recorder here, so somebody writes the schema. We do not defeat anti-bot systems or solve CAPTCHAs, which is part of what those Octoparse add-ons buy. There is no free tier. And on the narrow-and-deep workload above, six sites and half a million pages, Octoparse at $69 is simply the cheaper answer and we would be the wrong recommendation.

Where the page meter wins is the wide workload: many sources, modest depth, output that has to arrive as typed fields in a pipeline. If that is the shape of your job, our price monitoring API covers the catalog-tracking case that brings most people to Octoparse in the first place, and building lead lists from public sources covers the template use case. If you are comparing meters across the category more broadly, SerpApi pricing per 1,000 searches is the other unusual one worth understanding, and our web scraping API pricing breakdown puts every model on one page.