Search MCP server
{
"mcpServers": {
"neurobird-search": {
"type": "http",
"url": "https://search.neurobird.com/mcp"
}
}
}{
"mcpServers": {
"neurobird-search": {
"type": "http",
"url": "https://search.neurobird.com/mcp"
}
}
}{
"mcpServers": {
"neurobird-search": {
"type": "http",
"url": "https://search.neurobird.com/mcp"
}
}
}{
"servers": {
"neurobird-search": {
"type": "http",
"url": "https://search.neurobird.com/mcp"
}
}
}{
"mcpServers": {
"neurobird-search": {
"type": "http",
"url": "https://search.neurobird.com/mcp"
}
}
}file .mcp.json (in your project root)file claude_desktop_config.jsonfile ~/.cursor/mcp.jsonfile .vscode/mcp.jsonfile ~/.codeium/windsurf/mcp_config.jsonversion 1.0.0
What it does
A hosted MCP server that performs web searches and returns ranked results with relevant page passages already extracted, plus the ability to fetch and convert any URL to clean markdown. Designed for AI agents that need current web information without making multiple requests.
- Search the live web with a query and get ranked results with extracted passages to answer questions without follow-up page fetches.
- Fetch multiple URLs and extract their main content as markdown, with boilerplate and ads removed, optionally trimmed to passages relevant to a query.
- Perform specialized searches in restricted domains, time periods, or topics (news, code, science, finance) with configurable search depth.
Tools this server exposes
| Tool | Does |
|---|---|
| web_search | Search the live web and get ranked results with relevant passages extracted from each page, optionally including a grounded answer and with filters for domains, |
| extract_url | Fetch one or more URLs and return their main content as clean markdown or text, with navigation and boilerplate removed and PDFs converted to text. |
Before you enable it
An MCP server runs with whatever access you give it. Start with a development account rather than production, and check which tools the server exposes before you let an agent call them unattended.