Webpages in, structured records out.
Most ways of giving a webpage to an AI flatten it into a stream of text. Analog turns the page into the things it contains — products, articles, events, listings, and other records — so an AI can query, compare, export, and use them directly. No site-specific schema or API is required.
Analog is currently in private alpha. Join the private alpha →
Already have access? Install Analog · Run the quickstart · Use Analog through MCP
Pages are fetched in a browser on your machine by default. For structured
extraction, the SDK sends the URL and rendered page content to Analog. Results
return to the SDK and are saved locally unless disabled. Authenticated content
is sent only when you explicitly provide it through html= or a custom
fetcher. mode="local" is the Markdown-only path that sends nothing.
Extraction uses no LLMs. Given the same page URL and HTML, the same engine produces the same records. A live page's HTML can change between fetches.
analog-sdk— theanalogcommand, Python SDK, and built-in browser.analog-mcp— Analog for MCP-compatible agents.marcato— deterministic HTML-to-Markdown conversion onlxml.homebrew-tap— Homebrew distribution for the Analog CLI.
- Agentic delight. We love building things that AIs love: tools that are easy to use and transparent about their capabilities.
- Human outcomes. Behind every AI is a person trying to get something done. Analog needs to make things easier for people, not just AIs.
- Safety. We're building the interface between AI and the internet. AIs are obedient, and the internet is not always friendly. So we need to make sure that combination stays safe.
