
Screenshot.fyi
Screenshot.fyi captures pixel-perfect website screenshots through a single API call, and now your AI agent can trigger them mid-conversation. Customers ask to see a page, verify a design, or document a bug, and your agent delivers a high-quality capture instantly.




Custom viewports, device emulation, full-page scrolling, and format control. Your AI agent captures exactly what you need from any URL through Screenshot.fyi's rendering engine.
Screenshot.fyi
From design reviews to competitive analysis, see how AI agents use Screenshot.fyi to capture website visuals on demand, replacing manual screenshot workflows with instant automation.
A designer asks the AI agent to show how the new landing page renders on desktop, tablet, and mobile. The agent fires three Screenshot.fyi requests with different viewport sizes and device emulations, returning all three screenshots in the conversation. Responsive testing that used to require three devices happens in one chat message.
A customer reports that a page looks broken. The agent captures a Screenshot.fyi screenshot of the URL the customer mentions, at the viewport dimensions they describe. The visual evidence attaches to the support ticket automatically. Developers see the exact issue without asking the customer to manually screenshot and email.
Your marketing team wants to document a competitor's new campaign page before it changes. The agent captures a full-page screenshot through Screenshot.fyi, preserving every section from hero to footer. The visual archive captures the design at this exact moment in time, no manual scrolling or pasting required.

Screenshot.fyi
FAQs
Any custom viewport size you specify. The agent sets viewport_width and viewport_height in pixels. Common configurations include 1280x720 for standard desktop, 1920x1080 for full HD, 375x812 for iPhone, or 768x1024 for tablet. The defaults are 1280 wide by 720 tall if no custom values are provided.
Yes. Setting full_page to true tells Screenshot.fyi to scroll the entire page and capture everything from the top of the header to the bottom of the footer. The result is a single tall image containing the complete page content, regardless of the viewport height setting.
The delay parameter adds a wait time in milliseconds after the page loads but before the screenshot is taken. This gives JavaScript time to execute, animations to complete, and lazy-loaded images to appear. For pages with heavy async content, a delay of 2000-5000ms typically ensures everything renders.
No. Screenshots are generated in real-time by Screenshot.fyi and delivered directly in the conversation. Tars does not maintain a persistent archive of captured images. If you need to save screenshots, you can download them from the conversation or use the JSON output format for programmatic storage.
Yes. The agent passes a cookie string to Screenshot.fyi that gets set in the browser before loading the page. This is useful for capturing authenticated pages, testing different user states, or viewing content that varies based on session data. Standard cookie format like 'session=abc123; theme=dark' is supported.
Image output returns the raw screenshot binary, which displays directly in conversations. JSON output wraps the image in a base64-encoded string inside a JSON response. Use JSON when you need to process the image programmatically, store it in a database, or pass it to another API endpoint.
Screenshot.fyi focuses on simplicity and premium quality with a streamlined parameter set. ScreenshotOne offers animated captures in video and GIF formats with detailed scrolling scenarios. Choose Screenshot.fyi for clean static captures and ScreenshotOne when you need animated or video recordings of page interactions.
Yes. The agent sets a custom user_agent string that Screenshot.fyi uses for the request. This influences how the target site renders content, since many sites serve different layouts based on the user agent. Combine this with custom headers for complete request customization.
Don't limit your AI Agent to basic conversations. Watch how to configure and add powerful tools making your agent smarter and more functional.

Privacy & Security
At Tars, we take privacy and security very seriously. We are compliant with GDPR, ISO, SOC 2, and HIPAA.