AI-native Web Search MCP

Example Use Cases

• 1

  Cost-effective, aggregated web search results from multiple engines.

• 2

  Fast responses with high cache hit rates to reduce API costs.

• 3

  Flexible search categories and time filtering in their queries.

Description

search-mcp is a Rust-based MCP server that combines five search backends—Brave, Exa, SearXNG, Tavily, and DuckDuckGo—into one unified search tool. It aggressively caches results to reduce costs and latency, routes queries to the cheapest available backend first, and deduplicates and fuses results for better quality. This single binary solution requires no Python or Docker dependencies and supports batch queries, multiple categories, and time filtering.