Browse
→ Web & Browser
→ BetterWebSearch MCP
BetterWebSearch MCP
Keyless-first TypeScript MCP server for web research, using DuckDuckGo by default and optional Brave or Tavily backends. It combines web search with a three-tier extraction ladder from fast HTTP fetch to hydration and Playwright, plus cache and SSRF protections, so agents get cleaner research payloads without mandatory API keys. Public GitHub repository created August 29, 2026.
MCP unverified
Integration
| Transport | stdio |
| Auth | none |
| Endpoint | npx -y better-web-search-mcp |
| Install | npx -y better-web-search-mcp |
Use Cases
| 01 | Give MCP clients working web search immediately without configuring a paid search API key |
| 02 | Extract pages through fast HTTP, hydrated fetches, or Playwright when simple search snippets are not enough |
| 03 | Use optional Brave or Tavily credentials for stronger search backends while retaining keyless DuckDuckGo fallback |
Tags
web-search duckduckgo brave tavily playwright extraction ssrf-guard typescript
Machine-readable: /api/servers.json
· JSON-LD schema embedded in <head>