Skip to main content
Guides/Site Performance

Page Screenshot: Understand and Use the Results

Capture a web page as a JPEG screenshot

EdgeDNS Team·Updated ·4 min read

Page screenshots: a visual observation from one browser

A page screenshot records the pixels displayed by a browser at one point in time. A viewport screenshot captures the visible rectangle; a full-page screenshot attempts to capture a larger document, including content outside that rectangle. The viewport dimensions are part of the observation.

Screenshots help teams review a deployment, document an error page or show a rendering problem that is difficult to explain with status codes alone. A timestamped image can accompany logs and other measurements during an incident.

To interpret a capture, check:

  • The final URL and HTTP status so an error page is not mistaken for a successful deployment.

  • The viewport width and height, since responsive layouts change at different sizes.

  • The observation time and loading interval.

  • Whether cookies, login state or consent prompts affected the page.

An image can show that a page appeared a particular way in that browser. It does not establish accessibility compliance, interaction correctness or appearance on every device. A JPEG is compressed image data; base64 is a text encoding often used to carry that data inside JSON. Downloading the decoded JPEG avoids having to handle that text manually. A fresh screenshot and its metadata are most useful together.

The Page Screenshot endpoint, in plain language

Capture a web page as a JPEG screenshot

See what a public deployment displays and retain a visual snapshot for reviews or incident investigations. The playground previews the image and downloads it without requiring manual base64 decoding.

What is actually happening when you call it

Creates a fresh browser and context, loads the URL (web address) with JavaScript enabled, waits for DOMContentLoaded plus 500 milliseconds, and captures the 1280×720 viewport as a quality-70 JPEG. The screenshot is limited to 1 MiB and returned inline as base64; this is not a full-page image or a permanent image-hosting URL. Each response identifies the request environment and shared VPS browser runtime. Private destinations, custom scripts, credentials, service workers, downloads and WebSockets are blocked; TLS (Transport Layer Security) verification stays enabled. One browser job runs at a time; busy responses return 429 with Retry-After and are uncharged. No capture is served from cache.

Why this specific tool matters

Choose a task: Visual Deployment Check. Capture the public deployment after a release.

Run the Page Screenshot request below and retain the input and observed result. Before changing configuration, compare the relevant fields with the value or policy you expected. Use the Page Screenshot reference to check the field definitions and supported options.

Info:

Available on the developer plan. The technical details: `GET /v1/domain/screenshot`.

When would I actually use this?

Visual Deployment Check

Capture the public deployment after a release.

Use the finding: See obvious rendering failures alongside other EdgeDNS diagnostics.

Incident Evidence

Record the page shown from the VPS while investigating an outage.

Use the finding: Retain a timestamped visual observation from a known vantage point.

Review Downloads

Preview a public page in the playground and download its screenshot.

Use the finding: Share a JPEG without decoding the API (Application Programming Interface) response manually.

The easiest way: just ask your AI assistant

Connect the MCP server. Just ask in plain English:

Use domain_screenshot with url=https://example.com/. Show the returned findings, identify unavailable evidence, and explain a useful next check.

Confirm the tool and inputs in the client before accepting its summary. For Page Screenshot, use the response fields below to separate observations from recommendations.

Tip:

The dedicated MCP server is available on every plan. Individual tool access and request limits depend on your plan. Management categories and endpoints disabled for the tool catalog are not exposed through MCP.

The technical way: call it from code

This Page Screenshot request uses the example inputs from its API contract. Replace YOUR_API_KEY with a key on the developer plan or higher; adapt the target to your task.

bash
curl --request GET 'https://api.edgedns.dev/v1/domain/screenshot?url=https%3A%2F%2Fexample.com%2F' \
  --header "Authorization: Bearer YOUR_API_KEY"

Prepare inputs for Page Screenshot

Start with the inputs below. Keep the same hostname, path, and options when comparing Page Screenshot observations. The API reference contains the complete parameter contract, optional settings, and access requirements.

FieldTypeRequired?What it meansExample

url

string

Yes

Public HTTP (HyperText Transfer Protocol)(S) URL (web address) on port 80 or 443. Bare domains use HTTPS (secure HyperText Transfer Protocol). Credentials and fragments are not accepted.

https://example.com/

Interpret Page Screenshot findings

Read these fields alongside the full response from your Page Screenshot check. A missing value or failed lookup is different from a passing result. Keep returned errors and confidence or cache metadata when available; do not replace missing evidence with a successful default. See the complete response reference for additional fields.

FieldTypeWhat you'll see in it

screenshot.mime_type

string

Always image/jpeg.

screenshot.base64

string

Base64-encoded JPEG, decoded image at most 1048576 bytes.

screenshot.width

number

Viewport width: 1280 pixels.

Words you might be wondering about

API (Application Programming Interface) — A way for one program to ask another program for something — like a waiter taking your order to the kitchen.

URL (web address) — The full address of a page, like https://example.com/about.

TLS (Transport Layer Security) — The encryption that puts the 'S' in HTTPS. It scrambles data so nobody between you and a website can read it.

Need Programmatic Access?

Automate domain intelligence with 100+ API endpoints and a free MCP server for AI integration.

Page Screenshot: Results Guide | EdgeDNS