Search made
for documentation
DocSearch by Algolia makes your docs and blogs instantly searchable — fast, relevant, and AI-ready. Free for open-source and technical docs.
Already trusted by your favorite docs
Join 9,000+ projects finding answers in milliseconds
Built for developers
Solve docs challenges with a search engine
Docs are only helpful when your users can find answers easily. Enter DocSearch.
Powered by Algolia Agent Studio
Ask AI
Instant AI answers straight from your own docs. Users ask natural-language questions and get context-aware answers with real sources — with prompts and models you configure.
npx @docsearch/cli setup --cursorModel Context Protocol
MCP Server
Make your docs neural-searchable by any AI agent — Claude, Cursor, Codex, and more — over the Model Context Protocol, powered by Algolia neural search.
Made for docs
DocSearch’s crawler automatically indexes your docs on a schedule — no manual reindexing, no stale results. A visual config editor lets you tune exactly what gets picked up, and it understands code snippets, tables, markdown structure, and API references so your users get pinpoint answers every time.
Works everywhere
Drop DocSearch into any stack. First-class adapters for Docusaurus and VitePress, headless React components, and a vanilla-JS build for everything else — shipping both a command-palette modal and the Sidepanel.
Customizable
Tailor DocSearch to match your brand and UX needs—colors, fonts, layouts, and even search behaviors are under your control. Drop-in CSS variables and simple JS hooks make it effortless to blend search seamlessly into any docs site.
A11y
DocSearch follows WAI-ARIA best practices to ensure full keyboard, screen-reader, and voice-control support. Delight every user with an inclusive search experience that’s tested against WCAG 2.1 standards.
Powered by Algolia
Running on Algolia’s global search infrastructure — fast, reliable, and free for the docs that need it.
- <20 ms
- Search latency
- 99.99%
- Uptime SLA
- 9,000+
- Projects
- Free
- For OSS & technical docs
