What does JavaScript rendering really cost across scraping APIs?

The published answer runs from 1 request with rendering included to 12.66x for Zyte’s highest pay-as-you-go difficulty tier; common credit-metered APIs charge 5x or 10x, but several vendors use plan-specific rates instead. Olostep, ScrapingFish, Scraping Robot, SOAX, ScrapingBee, ScraperAPI, and Zyte, retrieved 2026-07-26.

Do not compare vendors by headline credits. This guide isolates the rendering change on the nearest ordinary proxy path. The scraping-credit guide owns the separate arithmetic for combining rendering with premium proxies or protected modes. Scrapfly billing and ZenRows pricing, retrieved 2026-07-26.

Every published JavaScript rendering meter in the current vendor records

“Factor” below compares the JavaScript path with the same vendor’s nearest ordinary HTTP path. A replacement rate is the whole request cost. An add-on is added to the base. A separate quota means there is no honest single multiplier, because both the included allowance and overage rate change.

Every published JavaScript rendering meter in the current vendor records table

One current-data conflict needs to stay visible. The Crawlbase target record says a rendered call consumes 2 requests. The vendor’s current Crawling API page says pricing is the same per successful response on the normal and JavaScript paths, then describes the normal token as “faster, cheaper.” This guide reports the published 1x statement but does not classify Crawlbase as verified free rendering. Crawlbase Crawling API, retrieved 2026-07-26.

Which vendors charge nothing extra for rendering?

Four current records have primary documentation supporting no rendering surcharge:

• Olostep defines one request as one webpage or PDF processed, says infrastructure costs are included, and supports pages that require JavaScript rendering. Olostep, retrieved 2026-07-26.

• ScrapingFish publishes one $0.002 successful-request rate and says its available features, including JavaScript execution, have the same price. ScrapingFish, retrieved 2026-07-26.

• Scraping Robot publishes a $0.0018 rate per successful scrape and includes JavaScript rendering in that rate. Scraping Robot, retrieved 2026-07-26.

• SOAX Web Data API says JavaScript rendering is included and publishes all-in rates of $0.90 to $1.30 per 1,000 requests by plan. SOAX Web Data API and SOAX pricing, retrieved 2026-07-26.

Crawlbase is excluded from that list. Its page contains both the same-price statement and the “cheaper” normal-token wording, while the current target record says 2 requests for rendering. The surcharge is therefore not verifiable from the available first-party documentation. Crawlbase Crawling API, retrieved 2026-07-26.

“Included with every plan” does not necessarily mean “no usage premium.” Oxylabs includes Headless Browser access in the subscription feature list but still publishes a higher per-1,000 result price when JavaScript is used. Oxylabs pricing, retrieved 2026-07-26.

Worked example: Scraping Robot at 120,000 successful scrapes

Scraping Robot’s public Business rate is $0.0018 per successful scrape from zero through 500,000 scrapes, with JavaScript rendering included. The worked input is 120,000 billable successful scrapes, of which 30% require rendering. The workload size and rendered share are hypothetical inputs, not vendor claims. Scraping Robot, retrieved 2026-07-26.

First split the workload, then apply the same dataset rate to both paths:

rendered successes = 120,000 × 30% = 36,000
ordinary successes = 120,000 - 36,000 = 84,000

rendered cost = 36,000 × $0.0018 = $64.80
ordinary cost = 84,000 × $0.0018 = $151.20
total = $64.80 + $151.20 = $216.00

The rendered share does not change this request bill because the current record has one successful-scrape rate and no separate rendering meter. Scraping Robot, retrieved 2026-07-26.

Worked example: Scraping Robot at 120,000 successful scrapes table

The surcharge attributable to switching this published workload from 0% rendering to 100% rendering is therefore $216.00 - $216.00 = $0.00. That does not claim the two paths complete the same amount of work in the same time; it states only that the published request rate is unchanged. Scraping Robot, retrieved 2026-07-26.

The record also defines where this arithmetic stops. At 500,000 successful scrapes, the published rate gives 500,000 × $0.0018 = $900.00. Above 500,000, the vendor advertises quote-dependent Enterprise pricing rather than a fixed public tier, so the Business rate should not be extrapolated past that boundary. Scraping Robot, retrieved 2026-07-26.

How to check whether you need rendering before paying for it

Test the required fields, not whether the page “uses JavaScript.” A site can run scripts while placing the data you need in the initial response; ScrapingFish specifically warns that content may be present in raw HTML even when a JavaScript-enabled browser changes what is displayed. ScrapingFish, retrieved 2026-07-26.

1. Fetch the target without a browser. Use a normal HTTP client, follow redirects, and inspect the response body:

   curl -L "$TARGET_URL"

   -L tells curl to follow HTTP redirects. curl manual, retrieved 2026-07-26.

2. Search that response for a field whose value you already know. Check the HTML, JSON-LD, and serialized application data. If every required field is present across representative URLs, parse that response and leave rendering off. Chrome DevTools Network panel, retrieved 2026-07-26.

3. If the field is missing, inspect the browser’s Network panel. Reload the page and compare the main Document response with Fetch/XHR responses. Chrome DevTools documents filters for Fetch/XHR and lets you inspect each response body. If a permitted, stable data endpoint returns the field directly, call that endpoint instead of paying to render the visual page. Chrome DevTools Network panel, retrieved 2026-07-26.

4. Enable rendering only for the URLs that fail the raw test. Crawlbase recommends using its normal token first and moving to the JavaScript token when the normal response is empty or challenge-blocked. Scrapfly likewise says to enable browser rendering only when the site requires it. Crawlbase Crawling API and Scrapfly billing, retrieved 2026-07-26.

5. Retest the cheap path when the target changes. ScrapingFish notes that desired content can be present in raw HTML even when opening the same response in a JavaScript-enabled browser changes what is displayed. ScrapingFish, retrieved 2026-07-26.

Pay for rendering when the required data appears only after client-side requests, interaction, or scrolling and there is no usable direct response path. Do not pay for it merely because the site is built with React, Vue, Angular, or another JavaScript framework. Crawlbase Crawling API and Scrapfly billing, retrieved 2026-07-26.

Four rendering meters added by the current dataset

Four current request.js_render records arrived after the earlier table was authored. Each publishes a replacement rendering mode rather than an additive rendering fee. The worked workload below is 10,000 requests. It converts each record’s basic and rendering meters into credits without treating one vendor’s credit as equal to another vendor’s credit.

Four rendering meters added by the current dataset table

The credit totals are the complete answer available from these four meter records. None publishes one universal cash value per credit that can be compared across vendors without selecting a plan or pack, so this section does not convert the totals into invented dollar prices. The linked records and sources were checked 2026-07-26.