home / mcp / openai web search mcp server
Provides real-time web search capabilities to MCP clients using OpenAI web_search_preview.
Configuration
View docs{
"mcpServers": {
"bitibi-openai-websearch-mcp-server": {
"command": "npx",
"args": [
"-y",
"openai-websearch-mcp-server"
],
"env": {
"OPENAI_API_KEY": "YOUR_API_KEY"
}
}
}
}You run an MCP server that adds real-time web search capabilities to your AI workflows. This OpenAI Web Search MCP Server enables models to perform live web lookups via OpenAI's web_search_preview feature, giving you current information from across the internet through the MCP protocol.
Install and run the server, then connect your MCP clients to access a web search tool named web_search. Use real-time searches to supplement model responses with up-to-date internet results. In practice, you can chain web_search with your prompts to gather sources, verify claims, or monitor breaking news, all within your MCP-enabled workflow.
Prerequisites: Node.js and npm must be installed on your system. You can verify by running node -v and npm -v.
Install or run the MCP server using either of the following options.
npx openai-websearch-mcp-server
```
```bash
npm install -g openai-websearch-mcp-serverSet up your environment and run the server locally. The server defaults to port 3000; you can override this with the PORT environment variable.
Environment variables are used to authorize with OpenAI. Example: set your API key in the environment before starting the server.
{
"mcpServers": {
"openai_websearch": {
"command": "npx",
"args": [
"-y",
"openai-websearch-mcp-server"
],
"env": {
"OPENAI_API_KEY": "your_api_key"
}
}
}
}Keep your OpenAI API key secure. Do not expose it in public code or shared configurations. Consider using environment-based secret management and rotation practices.
Web search tool that uses OpenAI's web_search_preview to fetch real-time results from the internet for your prompts.