Guides
How to give AI agents the web with ContextFetch: build research agents and RAG pipelines, turn any page into clean Markdown, crawl and map sites, and move over from other search APIs. Each guide has complete code in the language or framework it names, and we run that code against the live API before the guide is published, so the output you see on the page is what it really returned.
Build an agent end to end: research agents, RAG, assistants, on the stack you use.
One task, done right: search, fetch pages as Markdown, crawl, map, cite.
Move from another search or scraping API with the code you already have.
ContextFetch inside frameworks, no-code tools and MCP clients.
Small complete apps: monitors, trackers, digests, fact-checkers.
How agent search works: grounding, pruned context, robots, llms.txt.
Real errors and how to fix them.
Search and scraping APIs side by side, with claims you can check.