Manual Extraction

Master the extraction workflow and output formats.

Quick Start

Navigate to any web page
Click the Decant icon in the toolbar
Click Extract Content
Copy or Save your clean content

Output Formats

Markdown

Clean text with headings, links, and images. Perfect for Obsidian, Notion, or any note-taking app. Uses Mozilla Readability + Turndown under the hood.

JSON

Structured output with title, url, content, and metadata (word count, image count, tables, dates, emails). Ideal for automation and pipelines.

MCP

Model Context Protocol format. Includes source URL, word/token count, and AI-optimized structure. Best for feeding content to Claude, ChatGPT, or other LLMs.

Switch formats instantly with the toggle — no need to re-extract.

Extraction Options

Smart Extract

Uses Readability to find the main article content. Strips ads, navigation, and sidebars automatically. Disable for full page HTML when you need everything.

Include Images

Keeps image references in the output. Disable for text-only extraction, which saves tokens when feeding content to AI.

Detect Tables

Converts HTML tables to proper Markdown or JSON tables. Enable this when extracting pages with tabular data.

Full Page

Extracts the entire page body instead of the detected article. Use this for dashboards, wikis, and single-page applications where Readability cannot isolate a single article.

Copy & Save

Copy sends the extracted content to your clipboard. Save downloads it as a .md, .json, or .json (MCP) file.

Use keyboard shortcuts for even faster workflow.

Right-Click Menu

Extract selection — extracts only the selected text
Extract full page — extracts the entire page
Copy for AI — extracts and copies in MCP format

Keyboard Shortcuts

Action	Shortcut
Extract	`Ctrl` + `Shift` + `E`
Copy	`Ctrl` + `Shift` + `C`
Save	`Ctrl` + `Shift` + `S`

On macOS, use Cmd instead. Customize shortcuts at chrome://extensions/shortcuts.