mcp-omnisearch
by spences10·★ 308·Score 51
A MCP server providing unified access to multiple search engines and AI tools through a single interface.
Overview
mcp-omnisearch is a comprehensive MCP server that consolidates multiple search providers and AI tools into four main functions: web search, AI-powered answers, GitHub search, and content extraction. It supports providers like Tavily, Brave, Kagi, Exa, GitHub, Linkup, and Firecrawl. The server is flexible, allowing users to start with just one or two API keys and add more as needed, making it easy to get started while offering advanced capabilities for power users.
Try asking AI
After installing, here are 6 things you can ask your AI assistant:
When to choose this
Choose this server when you need unified access to multiple search engines and AI tools without switching between different interfaces.
When NOT to choose this
Don't choose this if you need specialized search functionality not supported by the integrated providers, or if you're concerned about vendor lock-in to multiple services.
Tools this server exposes
4 tools extracted from the READMEweb_searchSearch the web for information using multiple providers.
ai_searchGet AI-powered answers with citations and reasoning.
github_searchSearch GitHub for code, repositories, or users.
web_extractExtract, process, or summarize web content from URLs.
Comparable tools
Installation
{
"mcpServers": {
"mcp-omnisearch": {
"command": "node",
"args": ["/path/to/mcp-omnisearch/dist/index.js"],
"env": {
"TAVILY_API_KEY": "your-tavily-key",
"KAGI_API_KEY": "your-kagi-key",
"BRAVE_API_KEY": "your-brave-key",
"GITHUB_API_KEY": "your-github-key",
"EXA_API_KEY": "your-exa-key",
"LINKUP_API_KEY": "your-linkup-key",
"FIRECRAWL_API_KEY": "your-firecrawl-key",
"FIRECRAWL_BASE_URL": "http://localhost:3002"
}
}
}
}FAQ
- Do I need API keys for all providers?
- No, the server automatically detects which API keys are available and only enables those providers. You can start with just one or two keys and add more as needed.
- How do I handle large search results?
- You can configure the server with OMNISEARCH_LARGE_RESULT_MODE environment variable ('file' or 'inline'). Individual tools can also override this setting with a 'large_result_mode' parameter.
Compare mcp-omnisearch with
Last updated · Auto-generated from public README + GitHub signals.