Documentation
Keyless-first. Works out of the box with DuckDuckGo — no API keys. Add
BRAVE_API_KEYlater for richer ranking & recency.
A local Model Context Protocol server that gives AI agents genuinely better web research: multi-provider search, three-tier page extraction, and deep research — without locking you into a single search API.
At a Glance#
| Install | npx -y better-web-search-mcp |
| Keys required | None — DuckDuckGo works immediately |
| Node | 20+ |
| License | MIT |
| npm | better-web-search-mcp |
Tools#
| Tool | What it does | Keyless? |
|---|---|---|
web_search | Fast search (aggregated, deduped, reranked) | ✓ |
web_research / deep_search | Deep research: rewrites question → parallel searches → top-10 extraction → citations | ✓ |
web_extract | Clean extraction HTTP → hydration → browser with confidence & api_endpoints | ✓ |
web_find | site: scoped search | ✓ |
web_news | Recent news + timeline + diversity | ✓ |
Learn More#
- Quickstart — Get running in 30 seconds
- Installation — From source and npx
- Tools Reference — All 6 tools with input schemas
- Configuration — Environment variables
- Providers — Brave, Tavily, DuckDuckGo
- Extraction — 3-tier page extraction pipeline
- Architecture — System design overview
- Changelog — Release history
- Contributing — How to help