Search the web in real-time through an MCP server. For developers building AI agents and assistants that need current information, news, or fact-checking capabilities.
io.github.bytedance/mcp-server-search
Local install
STDIO
No auth required
How models use it and what it is built for.
Search the web in real-time through an MCP server. For developers building AI agents and assistants that need current information, news, or fact-checking capabilities.
Local install — runs as a subprocess.
Where to find authoritative docs and source for Bytedance MCP Server Search.
Paste any of these into Agent Studio after connecting Bytedance MCP Server Search.
Common questions about connecting and running Bytedance MCP Server Search.
What search engine does this MCP server use?
The registry metadata does not specify which search engine backend powers this server. Check the full documentation or source repository for details on the search provider and any API dependencies.
How do I install and run the search MCP server?
Install via npm with `npx @agent-infra/mcp-server-search@latest`. The server runs over stdio transport, so it integrates directly into your MCP client without additional network configuration.
Can I use this server offline or does it require internet?
This is a web search server, so it requires internet connectivity to fetch real-time results. It is designed for live queries, not local or cached data.
Are there rate limits or usage quotas?
The registry metadata does not document rate limits or pricing. Review the package documentation or source code for any usage restrictions or authentication requirements.
How does this compare to building search directly into my agent?
Using an MCP server abstracts search as a reusable tool your agent can call, making it easier to swap providers, manage dependencies, and share across multiple projects without duplicating code.
MCP Playground runs 10,000+ hosted MCP servers — GitHub, Linear, Notion, Stripe, Sentry and more — across Claude, GPT, Gemini, DeepSeek and 30+ AI models. Compare model answers side-by-side, save agent presets, share runs. Zero install.
Open Agent Studio