Screenshot API Options

The URL you would like to capture. For example:

https://wikipedia.com

Emulate how a web page behaves and displays content in a set time zone. The value for the time_zone field must be a supported ICU programmatic timezone ID (e.g. "US/Eastern", "Europe/Berlin", "Asia/Tokyo", "America/Sao_Paulo", "Pacific/Auckland"). You can find the latest full list of timezone IDs in this Unicode txt document, or for a quicker reference, see this Wikipedia article.

Capture only a specific element on a web page by providing a CSS Selector that identifies the element. Note: Accepts a single CSS Selector.

Requires selectors field. Displays available :hover styles on the element identified by the selectors field when set to true.

See details and an example in this section.

Set the real_location parameter to a supported city, country, or US state to take screenshots from that geographical location through a proxy network.

The id of a proxy stored on your account (see PUT /proxies). The capture is routed through that proxy instead of the PagePixels network, so no Real Location bandwidth is used. Requires the Bring Your Own Proxy addon. Replaces real_location when both are given.

Provide a CSS selector that should appear on the page before the screenshot is captured. For example:

#id-to-wait-for

Slowly scrolls to trigger content that only loads when a user scrolls (lazy-loaded). Note: May take up to 10 seconds to complete.

Captures the entire web page using a more advanced algorithm. Useful for pages with lazy-loaded content or when the standard fullpage option doesn’t produce the desired result. Note: This option takes a longer time to process vs fullpage.

Width of the viewport when the screenshot is taken. Default is 1920 (pixels). Use this option to emulate different mobile and desktop device dimensions.

Height of the viewport when the screenshot is taken. Default is 1000 (pixels). Use this option to emulate different mobile and desktop device dimensions.

The image at the provided URL will be displayed until the first screenshot is captured. The provided URL should return an image.

Quality defines the JPEG compression, where 1 produces the lowest quality image and 100 creates the highest quality image. The higher the quality, the larger the image will be. This setting is ignored if you are producing a PNG or WebP image.

• "Extract the business's current rating from this page."
• "Analyze the branding elements present in the image."
• "Identify the target audience based on the marketing material."
• "Summarize the key marketing message conveyed in the image."

Which AI provider analyzes the screenshot. pagepixels (the default) uses your PagePixels AI Analysis credits. anthropic, openai and openrouter send the analysis to your own provider account using the API key you have stored with PUT /ai/providers/:provider, and require the Bring Your Own AI addon. Your PagePixels credits are never consumed when using your own provider.

The model id to use with your own provider. Required for openai and openrouter; optional for anthropic. Only vision-capable models that support structured outputs are accepted — list them with GET /ai/models?provider=…. Ignored when ai_provider is pagepixels.

simple (the default) sends ai_prompt and returns the answer as free text in ai.ai_response. structured ignores ai_prompt and instead returns a JSON object matching ai_structured_fields, available as ai.ai_structured_response (and JSON-encoded in ai.ai_response).

Optional context for a structured analysis, for example This is an e-commerce product page. Only used when ai_prompt_mode is structured.

The fields to extract, as a JSON array (or a JSON string on GET requests). Each field has a name (letters, numbers, underscores), a type of string, number, boolean, enum (pick one) or multi (pick several), an optional description, and for enum/multi an options array. Up to 50 fields, up to 100 options each. Example:

[{"name":"page_loaded","type":"boolean","description":"Did the page render without errors?"},{"name":"status","type":"enum","options":["ok","warning","broken"]},{"name":"issues","type":"multi","options":["layout broken","missing images","error message"]}]

Add additional HTTP headers to your request.

Add custom cookies to your request.

Instruct the browser to ask the website to display a specific language. This is great for websites that support multiple languages.

Learn More about Accept Language

Give user device information to the browser. This can be useful for emulating different devices.

Learn More about User Agents

Provide latitude location information to the browser.

Learn More about Latitude

Provide longitude location information to the browser.

Learn More about Longitude

Provide accuracy location information to the browser.

Learn More about Accuracy

Ads are not blocked by default. To block Ads, set this option to true.

Tracking is not blocked by default. To stop tracking scripts from running, set this option to true.

Cookie consent banners are not removed by default. To hide cookie consent banners, set this option to true.

Javascript is enabled by default. To disable all javascript, set this option to true.

To discard Javascript resources from outside of the requested domain, set this option to true. This can help speed up load times, improve performance, and reduce the risk of a timeout.

Add a custom title to your screenshot configuration

Add a custom description to your screenshot configuration.