How to automate screenshots of Indian marketplace product pages with Urlbox
Automate screenshots of Amazon.in and Flipkart product pages with Urlbox render links or its API, while accounting for dynamic content, access rules, and capture failures.
To automate screenshots of Amazon.in or Flipkart product pages with Urlbox, submit each authorized product-page URL as a render request, select a viewport or full-page capture, then save and inspect the returned image. Use Urlbox render links for a simple integration or its JSON API for application workflows; waits and selectors can help with dynamic pages. Urlbox’s documentation is generic: it does not verify consistent rendering of either marketplace, provide marketplace-specific selectors, or guarantee access. Pilot your exact URLs and use case before scheduling captures broadly.
This guide covers a customer-initiated screenshot workflow. It is not a method for crawling marketplaces, bypassing access controls, or establishing permission to capture a page. Check the applicable marketplace terms and access rules before you automate.
1. Prepare the URL list and capture requirements
Treat each product page you intend to capture as a separate render job. Urlbox describes its service as rendering customer-submitted URLs rather than following links or spidering a site. Keep the input URL, requested capture settings, capture time, and result together in your own workflow so you can trace an image back to its source.
Before implementation, decide:
- What to capture: the initial viewport, the full scrollable page, or a specific product detail element.
- Which viewport: set dimensions that match the device or audit requirement. Marketplace layouts and responsive breakpoints may change what is visible.
- Which output: choose an image format for an image archive or report; use PDF if a paginated document is more useful.
- When a page is ready: identify a page element to wait for, or a loading indicator that should disappear. Do not assume that a fixed delay means product data has finished loading.
- What overlays should appear: cookie prompts, promotions, and other overlays can be part of the record. Hide or dismiss them only when that matches your capture purpose.
2. Choose Urlbox render links or its API
Urlbox documents two integration patterns: render links that can be embedded directly, and JSON API requests that can be made synchronously or asynchronously. Its overview also points to batch processing and webhook workflows. Consult the Urlbox documentation for current request authentication, parameter syntax, plan availability, and response handling before wiring a production job.
A render link is useful when a URL can be assembled where a link is needed. An API integration is generally a better fit when your application must queue product URLs, store results, retry failures, or coordinate asynchronous jobs. In either case, the requested URL is the capture target; the workflow does not imply that Urlbox will discover more pages for you.
3. Pick viewport, full-page, or element capture
Viewport screenshot
A normal viewport capture records the visible browser area at the dimensions you request. It is a sensible starting point for a product card or above-the-fold snapshot. It will not include details lower down the page unless they happen to be visible in that viewport.
Full-page screenshot
Urlbox documents two full-page approaches:
| Mode | Behavior | Trade-off |
|---|---|---|
| Stitch mode | Scrolls through the page to trigger animations and lazy-loaded content, freezes fixed or sticky elements, captures sections, and stitches them together. | Described as the more reliable mode and optimized for accuracy. Long or unusual layouts should still be inspected. |
| Native mode | Uses the browser’s native full-page screenshot function. | Described as faster, but it may not work well on every site; optimized for speed over accuracy. |
For long marketplace pages, choose based on whether completeness or speed matters more, then inspect representative output. Lazy-loaded product images may not be present until their region is reached. Stitch scrolling is intended to trigger such content, but it is not a guarantee that every page element will load.
CSS-selector screenshot
A CSS selector can target a particular element, such as a product detail or price block, instead of capturing the whole page. Urlbox can be configured to fail when the requested selector is not found. This is useful for detecting layout changes rather than silently saving an unrelated region. The reviewed docs do not supply selectors for Amazon.in or Flipkart; identify and validate one against the actual page layout and locale you use.
4. Handle dynamic loading and overlays
Use a wait condition that reflects the page behavior. Urlbox documents wait_for for an element that should appear and wait_to_leave for a loading indicator that should disappear. Stitch mode can scroll to trigger lazy loading. The exact selectors and timing depend on the page, and the generic documentation does not establish marketplace-specific loading behavior.
Urlbox also documents options to hide cookie banners, attempt to click an accept button, press Escape, and block ads or selected URLs. These controls alter the captured page. Use them only when appropriate to your purpose and applicable consent requirements; a capture with an overlay dismissed may not represent what a visitor saw.
Do not treat stealth or headless-browser options as a permission mechanism. Urlbox documents use_stealth and hide_headless, and some options may depend on the plan. Their existence does not establish permission to bypass a marketplace’s access controls or guarantee a successful capture.
5. Build a repeatable capture workflow
- Collect only the product URLs and locales your use case is allowed to capture.
- Choose a render-link or JSON API integration and consult Urlbox’s current docs for authentication and exact parameter names.
- Set the viewport, output format, and capture mode. Start with a viewport or selector capture if a full page is not needed.
- Add an element wait or loading-indicator wait when the page needs it. Use the shortest condition that reliably represents readiness in your pilot.
- Submit each URL as an individual job, or use the batch and scheduled workflow appropriate to your integration.
- Save the response image with the original URL, settings, and capture time. For asynchronous jobs, use the documented webhook workflow to learn when a job completes.
- Inspect output images and failures, especially after page layout changes. Handle errors and retries explicitly rather than assuming every request produced a usable image.
Urlbox documents synchronous and asynchronous API requests and points to batch and webhook guides. Exact request fields, authentication, and plan requirements can change; use its current documentation for a runnable Urlbox request rather than copying guessed endpoint syntax.
6. Access rules and marketplace reliability
Urlbox’s bot disclosure says: “We are not a search or training crawler. We do not follow links or spider a site.” It also says signed requests honor robots.txt for the Urlbox product token, falling back to the * group. This describes Urlbox’s request behavior. It does not decide whether your marketplace use is permitted, and it does not promise that a particular page will render.
The reviewed Urlbox sources do not provide a verified compatibility matrix for Amazon.in or Flipkart, a tested success rate, an approved marketplace API route, or current marketplace-specific rules. Validate your exact URLs, locale, viewport, schedule, waits, and capture mode. Do not infer that an automation-related option ensures access or that an initial successful capture will remain reliable indefinitely.
7. Performance, reliability, and cost considerations
Capture choice affects the work involved: Urlbox characterizes native full-page capture as faster and stitch mode as more accuracy-oriented. A selector capture can reduce the intended image area, while a full-page capture can involve scrolling and assembling sections. The sources do not provide numerical performance figures or marketplace-specific success rates, so benchmark your own authorized pilot if throughput matters.
For reliability, retain enough information to diagnose a bad or missing result: source URL, timestamp, capture settings, job status, and the returned image or error. Use asynchronous jobs and webhooks where that fits the volume and latency of your workflow. Check Urlbox’s current plan and pricing pages for the cost of your expected capture volume and any plan-dependent options; the reviewed material does not establish prices.
Set a bounded schedule and inspect samples. A page can change its layout, loading behavior, or overlay state. Avoid launching a large run until a small pilot demonstrates that the selected output is useful and the access pattern is appropriate.
8. Troubleshooting
| Symptom | Likely cause | What to do |
|---|---|---|
| Blank or incomplete image | The requested page content had not appeared, or the capture mode did not suit the layout. | Check the target URL and response status; add a relevant wait_for condition or use wait_to_leave for a loading marker. Compare viewport and stitch modes. |
| Product images are missing below the fold | Images may load lazily as the page is scrolled. | Try stitch full-page mode, which scrolls to trigger lazy-loaded elements, then inspect the resulting image. |
| Selector capture fails | The selector is absent, changed, or not yet present when the capture runs. | Inspect the current page layout, update the selector, wait for it to appear, and keep failure-on-missing-selector behavior if you need layout changes to surface as errors. |
| Cookie or promotional overlay obscures content | The page is showing a consent banner or another overlay. | Decide whether the overlay belongs in the record. If not, use a documented hide or dismiss option where appropriate and verify what the result contains. |
| Capture is refused or the page does not render | The target may restrict automated access, or the URL, locale, or request conditions may not work. | Review the applicable marketplace rules and Urlbox’s access behavior. Do not treat stealth options as permission or a guarantee; use an authorized capture route if one is available. |
| Full-page result is distorted or unexpectedly long | Sticky elements, animations, or unusual page layout can complicate full-page capture. | Inspect stitch and native modes on a small sample; choose the one that preserves the content needed for your use case. |
| Scheduled jobs fail after previously working | The page layout, selector, or loading behavior may have changed. | Record failures, review the latest output, and revalidate waits and selectors before resuming a broad schedule. |
9. Or skip the browser setup
ScreenshotNeo provides a website screenshot API and MCP server. A single GET request takes a URL and returns an image or PDF. For a product page you are authorized to capture, this cURL example saves a WebP screenshot:
curl -G "https://api.screenshotneo.com/v1/shot" -d access_key=YOUR_API_KEY --data-urlencode url=https://stripe.com -o shot.webpReplace the example target with your authorized product-page URL. See the ScreenshotNeo API documentation for request options and setup.
ScreenshotNeo removes cookie banners, popups, and chat widgets before the shot. Bot checks, blank pages, and failed loads are never billed. Its MCP server lets AI agents take screenshots. The free plan includes 1,000 screenshots a month with no card; paid plans start at $5 for 3,000. Learn about ScreenshotNeo, then sign up for 1,000 free screenshots a month with no card.
FAQ
Can I schedule screenshots of Amazon.in or Flipkart pages?
Urlbox describes scheduled screenshot workflows, but the reviewed sources do not verify marketplace-specific rendering or permission. Confirm the intended use is allowed and pilot the exact pages before scheduling.
Does Urlbox crawl product listings automatically?
Urlbox says it renders customer-submitted URLs and does not follow links or spider a site. Supply the URLs your workflow is intended to capture.
Which full-page mode should I use?
Use stitch mode when accuracy and triggering lazy-loaded content are priorities; consider native mode when speed matters and verify the output on your page.
Are Amazon.in or Flipkart CSS selectors provided?
The reviewed documentation does not provide marketplace-specific selectors. Validate selectors against your target page and locale.
Does a stealth option guarantee a screenshot?
No. The existence of automation-related options does not guarantee a render or establish permission to bypass access controls.
