Website/Webpage URL Screenshot

What this tool does

Enter a public web address, choose a viewport, and the server asks a headless capture service for a screenshot of that page. The image comes back to the page in a preview panel, and you can download it as PNG, JPG, WEBP, BMP or GIF. Nothing is installed in your browser; the rendering happens on the server side, so the result does not depend on your screen size, your browser extensions or whether your laptop is awake.

It captures the viewport, not the entire scrolling page. A long article will be cut at the height you selected, and there is no full-page option in this interface. If you need the whole page, capture at a tall custom height within the allowed range, or take several screenshots at different scroll positions with a dedicated tool.

How the capture runs

1. The URL is validated: it must be a complete http or https address.
2. Private and internal hosts are rejected, so the tool cannot be used to probe a local network or a server behind a firewall.
3. The server sends the address to the screenshot service with the requested viewport width and height and asks for an image response.
4. The returned PNG is saved, converted to your chosen format if necessary, and cropped if you supplied crop dimensions.
5. The response includes a public URL for the image, which the page loads into the preview and reuses for the Download button.

Because a third-party renderer loads the page, what you get is that renderer's view: its IP address, its location, its cookie jar and its default user agent. A site that personalises content by country, or blocks data-centre traffic, may return something different from what you see in your own browser.

Viewport options

The Viewport selector offers common desktop sizes and a few phone sizes:

• Desktop: 1280×720, 1024×768, 1280×800, 1366×768, 1920×1080 (the default).
• Mobile: iPhone 14 (390×844), iPhone 16 (393×852), Galaxy (360×780), Pixel (412×915).
• Custom Viewport: enter any width from 320 to 5000 pixels and height from 240 to 10000 pixels.

Choosing a preset clears the custom width and height fields, so the preset always wins when one is selected. The preset only sets the viewport dimensions; whether a responsive site serves its mobile layout depends on the capture service and on how the site detects devices.

Formats and quality

PNG is the native output and the safest choice for screenshots with text, sharp edges or transparency. JPG produces smaller files for photographic pages. WEBP is smaller still and widely supported in browsers, BMP is uncompressed and large, and GIF is limited to 256 colours, so it is mainly useful for simple graphics rather than text-heavy pages. The capture service returns a PNG; the server converts it when you pick another format, and conversion can soften very fine text slightly.

The Quality field runs from 0 to 100 and defaults to 90. It matters for lossy formats such as JPG and WEBP; PNG, BMP and GIF ignore it. If a page looks fuzzy at quality 50, raise it rather than blaming the renderer.

Cropping

Crop Width and Crop Height trim the image from the centre. There is no offset control, so you cannot choose a top-left region; the crop is always centred on the captured viewport. Both fields must be filled for a crop to happen. If either is empty, the screenshot is returned at full size.

Crop Unit decides how the numbers are read. In pixels, a 800×450 crop takes an 800 by 450 region from the middle. In percent, 50×50 takes half the width and half the height, and values are capped at 100 percent. A crop larger than the source is clamped to the source size, so asking for 9000 pixels from a 1280-pixel-wide capture simply returns the full width. Cropping is done with the server's image library and re-encodes the file, which is why the quality setting also affects a cropped JPG or WEBP.

Using the page

1. Paste the URL, including https://.
2. Pick a format, a viewport and, if you want one, a crop.
3. Press Capture and wait. A slower page can take many seconds; the button stays disabled until the request finishes.
4. Inspect the preview. The zoom slider scales the image in the panel without changing the file, and the reset button returns it to 100 percent.
5. Press Download to save the same image the preview shows. The file name comes from the server, in the form capture-YYYYMMDD-HHMMSS-random.ext.

The status line reports success or the error text from the capture service. Errors are often the service explaining why it could not load the page, such as a timeout or a blocked request, rather than a problem with your input.

What changes the result

• Cookie banners and modals appear because the renderer is a fresh visitor; they can cover the content you wanted.
• Lazy-loaded images may be missing if they only load on scroll, since only the initial viewport is rendered.
• Login-gated pages show the sign-in screen, because the capture is anonymous.
• Animations are captured at whatever frame the renderer reaches, so a carousel may be mid-transition.
• Geo-targeted content reflects the renderer's location, not yours.
• Fonts may fall back if the renderer cannot fetch a web font quickly.

Limits

Only public HTTP and HTTPS addresses are accepted; localhost, private IP ranges and internal hostnames are blocked on purpose. The viewport is capped at 5000 pixels wide and 10000 pixels tall, and the crop at the same bounds. Captures are cached for three minutes per URL and option set, so pressing Capture twice with identical settings returns the same stored image instead of a fresh render. There is no batch mode, no scheduled capture, no authentication, no cookie injection and no way to scroll before capturing.

The Timeout field accepts 3 to 120 seconds, but the outbound request uses the capture service's configured timeout rather than that field, so changing it may not change how long the server waits. If a capture fails at the service side, retrying later is usually more effective than raising the number.

Privacy and where the image goes

The page is fetched by a third-party rendering service, which means the target site sees that service's request rather than yours. The resulting image is written to a temporary public folder and returned as a URL, so anyone with the link could open it until the folder is cleaned. Do not capture private dashboards, admin panels, invoices or anything containing personal data. The URL you submit is also sent to the capture provider.

Common mistakes

• Forgetting the scheme. example.com fails validation; https://example.com works.
• Expecting a full-page shot. Only the viewport is captured.
• Filling one crop field. Both width and height are required for a crop.
• Wanting an off-centre crop. The crop is always centred.
• Capturing a private URL. Internal hosts are rejected, and public temp links are shareable.
• Assuming a repeat capture is fresh. Identical settings are cached for three minutes.

Frequently Asked Questions