Browser Automation

Extract content from web pages, capture screenshots, or execute complex browser tasks using AI — all through a single API with no account setup required.

Ways to call this

In a workflow

Browser fetch/screenshot and sessions aren’t in the typed ctx.sapiom.* client — call their REST APIs with the SDK (@sapiom/fetch) from step code, as in the Quick Example. See Using Capabilities.

Direct (remote MCP)

Call sapiom_fetch, sapiom_screenshot, or sapiom_browser_session_create directly via the remote MCP — no workflow to author. Run tool_discover to find tools by goal.

SDK / REST

Wrap your HTTP client with @sapiom/fetch (or @sapiom/axios) and call the gateway directly — from anywhere, including inside a workflow step. See the Quick Example below and the API reference for full parameters.

Quick Example

import { createFetch } from "@sapiom/fetch";

// Create a Sapiom-tracked fetch function
const sapiomFetch = createFetch({
  apiKey: process.env.SAPIOM_API_KEY,
  agentName: "my-agent",
});

// Extract webpage content - SDK handles payment/auth automatically
const response = await sapiomFetch(
  "https://anchor-browser.services.sapiom.ai/v1/tools/fetch-webpage",
  {
    method: "POST",
    headers: { "Content-Type": "application/json" },
    body: JSON.stringify({
      url: "https://example.com",
      format: "markdown",
    }),
  }
);

const data = await response.json();
console.log("Page content:", data.content);

How It Works

Sapiom routes browser automation requests to Anchor Browser, which provides AI-powered browser automation in the cloud. The SDK handles payment negotiation automatically — you pay based on the operation type and complexity.

The service supports two operations:

1. Extract — Get page content as markdown or HTML
2. Screenshot — Capture page screenshots at various sizes

Provider

Powered by Anchor Browser. Anchor provides headless browser infrastructure with AI capabilities for intelligent web automation.

API Reference

Extract Content

Endpoint: POST https://anchor-browser.services.sapiom.ai/v1/tools/fetch-webpage

Extract the main content from a webpage as clean markdown or HTML.

Request

Parameter	Type	Required	Description
url	string	Yes	URL to extract content from
format	string	No	Output format: markdown or html (default: markdown)
wait	number	No	Wait time in milliseconds for JavaScript execution
new_page	boolean	No	Open URL in a new browser page (default: false)
page_index	number	No	Target page index in multi-page session
return_partial_on_timeout	boolean	No	Return partial content if page times out (default: false)

{
  "url": "https://example.com/blog/article-title",
  "format": "markdown"
}

Response

{
  "content": "# Article Title\n\nThis is the main content of the article...",
  "title": "Article Title",
  "url": "https://example.com/blog/article-title"
}

Capture Screenshot

Endpoint: POST https://anchor-browser.services.sapiom.ai/v1/tools/screenshot

Capture a screenshot of a webpage.

Request

Parameter	Type	Required	Description
url	string	Yes	URL to screenshot
width	number	No	Viewport width in pixels, 320-3840 (default: 1280)
height	number	No	Viewport height in pixels, 240-2160 (default: 720)
image_quality	number	No	JPEG quality, 1-100 (default: 80)
wait	number	No	Wait time in milliseconds for JavaScript execution
scroll_all_content	boolean	No	Scroll through page to capture all content (default: false)
capture_full_height	boolean	No	Capture full page height, not just viewport (default: false)
format	string	No	Image format: png or jpeg (default: png)
s3_target_address	string	No	S3 URL for direct upload instead of returning image
fullPage	boolean	No	Deprecated — use capture_full_height + scroll_all_content instead
quality	number	No	Deprecated — use image_quality instead

{
  "url": "https://example.com",
  "capture_full_height": false,
  "format": "png",
  "width": 1280,
  "height": 720
}

Response

The response is binary image data with the appropriate Content-Type header (image/png or image/jpeg). Use responseType: "arraybuffer" (Axios) or response.arrayBuffer() (Fetch) to handle the binary response.

Price Estimation

Endpoints:

• POST https://anchor-browser.services.sapiom.ai/v1/tools/fetch-webpage/price
• POST https://anchor-browser.services.sapiom.ai/v1/tools/screenshot/price

Get the estimated cost before making a request. Accepts the same parameters as the main endpoint.

{
  "price": "$0.02",
  "currency": "USD"
}

Error Codes

Code	Description
400	Invalid request — check URL and parameters
402	Payment required — ensure you’re using the Sapiom SDK
404	Page not found or unreachable
429	Rate limit exceeded

Browser Sessions

Long-lived Anchor browser sessions let you connect a CDP-compatible client (e.g. Playwright) to a cloud browser, then terminate and settle the charge at actual usage. These operations are reached via the Sapiom gateway/SDK (@sapiom/fetch) — there is no typed ctx.sapiom method for sessions.

Operation	Description
sapiom_browser_session_create	Open a paid Anchor browser session; returns sessionId, cdpUrl, and expiry metadata. Connect Playwright via chromium.connectOverCDP(cdpUrl).
sapiom_browser_session_create_with_identity	Open a session with a stored identity attached — Anchor restores the identity’s browser profile (and auto-logs in when credentials are present). Requires identityId (UUID from sapiom_browser_identity_create).
sapiom_browser_session_terminate	Close a session and settle the X402 authorization at actual duration cost. Always call this when done — open sessions hold the authorization until idle timeout (~5 min). Requires sessionId.
sapiom_browser_identity_create	Create an Anchor identity (free, no X402 charge) that stores browser profile and/or credentials. Returns an id — store it, as identities are not listable by name. Requires source (login page URL) and credentials array.

All four operations accept an optional agentName parameter for spend attribution and rule scoping.

Complete Example

import { createFetch } from "@sapiom/fetch";

const sapiomFetch = createFetch({
  apiKey: process.env.SAPIOM_API_KEY,
  agentName: "my-agent",
});

const baseUrl = "https://anchor-browser.services.sapiom.ai/v1";

async function scrapeArticle(url: string) {
  // Extract article content as markdown
  const response = await sapiomFetch(`${baseUrl}/tools/fetch-webpage`, {
    method: "POST",
    headers: { "Content-Type": "application/json" },
    body: JSON.stringify({
      url,
      format: "markdown",
    }),
  });

  const data = await response.json();
  return {
    title: data.title,
    content: data.content,
  };
}

async function capturePagePreview(url: string) {
  // Capture a screenshot for social preview
  const response = await sapiomFetch(`${baseUrl}/tools/screenshot`, {
    method: "POST",
    headers: { "Content-Type": "application/json" },
    body: JSON.stringify({
      url,
      width: 1200,
      height: 630,
      format: "jpeg",
      image_quality: 90,
    }),
  });

  return Buffer.from(await response.arrayBuffer());
}

// Usage
const article = await scrapeArticle("https://blog.example.com/post");
console.log("Article:", article.title);

const preview = await capturePagePreview("https://example.com");
console.log("Preview image size:", preview.byteLength, "bytes");

Pricing

Operation	Price
Extract	$0.01 flat
Screenshot	$0.01 flat