OpenDeepSearch MCP Server

This is a Model Context Protocol (MCP) server for OpenDeepSearch that allows LLM applications to interact with OpenDeepSearch's search capabilities.

Features

Exposes OpenDeepSearch's search functionality as MCP tools
Integrates with Claude Desktop and other MCP-compatible clients
Provides a standardized interface for LLM applications to access web search capabilities

Setup

This project uses uv for dependency management.

Install uv: Follow the instructions here.
Sync Dependencies: Navigate to the mcp_server directory and run: bash uv sync This will install dependencies based on pyproject.toml and uv.lock.

Configuration

The server requires certain environment variables to function correctly, especially API keys for the underlying services. These can be set directly in your environment or passed via the MCP client configuration (e.g., using Smithery CLI).

Variable	Description	Required	Default	Notes
LLM Providers	(Provide at least one)
`OPENAI_API_KEY`	API key for OpenAI LLM.	Optional	None	Needed if using OpenAI models.
`OPENAI_BASE_URL`	Custom base URL for OpenAI compatible endpoints.	Optional	None
`ANTHROPIC_API_KEY`	API key for Anthropic LLM.	Optional	None	Needed if using Anthropic models.
`OPENROUTER_API_KEY`	API key for OpenRouter.	Optional	None	Needed if using OpenRouter models.
`FIREWORKS_API_KEY`	API key for Fireworks AI.	Optional	None	Needed if using Fireworks models.
`GEMINI_API_KEY`	API key for Google Gemini.	Optional	None	Needed if using Gemini models.
`AZURE_API_KEY`	API key for Azure OpenAI Service.	Optional	None	Needed if using Azure OpenAI models.
`AZURE_API_BASE`	API base URL for Azure OpenAI Service.	Optional	None	Needed if using Azure OpenAI models.
`AZURE_API_VERSION`	API version for Azure OpenAI Service.	Optional	None	Needed if using Azure OpenAI models.
`AZURE_DEPLOYMENT_ID`	Deployment ID for Azure OpenAI Service.	Optional	None	Needed if using Azure OpenAI models.
`DEEPSEEK_API_KEY`	API key for DeepSeek.	Optional	None	Needed if using DeepSeek models.
Search Providers
`SERPER_API_KEY`	API key for Serper search provider.	Optional	None	Required if `search_provider` is set to `'serper'` (either by default or via tool argument).
`SEARXNG_INSTANCE_URL`	URL of your SearXNG instance.	Optional	None	Required if `search_provider` is set to `'searxng'` (either by default or via tool argument).
`SEARXNG_API_KEY`	API key for your SearXNG instance (if required by the instance).	Optional	None	Used if `search_provider` is set to `'searxng'`.
Rerankers
`JINA_API_KEY`	API key for Jina AI Reranker.	Optional	None	Required if `reranker` is set to `'jina'` (either by default or via tool argument).
Other Tools
`WOLFRAM_ALPHA_APP_ID`	App ID for WolframAlpha tool integration (if enabled in the agent).	Optional	None
Server Behavior
`LOG_LEVEL`	Controls the server's logging verbosity (DEBUG, INFO, WARNING, ERROR, CRITICAL).	Optional	INFO	Can also be set via the `--log-level` CLI argument passed by `smithery.yaml`.

Note: API keys passed directly as arguments to the perform_search tool (serper_api_key, searxng_api_key, jina_api_key) will temporarily override the environment variables for that specific call.

Usage with Smithery CLI

You can run this server using the Smithery CLI and the provided smithery.yaml configuration file. This allows you to easily manage the required environment variables.

# Example: Run with OpenRouter key and Serper key
npx -y @smithery/cli@latest run . --config '{"openrouterApiKey":"sk-or-...", "serperApiKey":"your-serper-key"}'

# Example: Run with OpenAI key and SearXNG
npx -y @smithery/cli@latest run . --config '{"openaiApiKey":"sk-...", "searxngInstanceUrl":"https://your-searxng-instance.com"}'

# Example: Run with Gemini key
npx -y @smithery/cli@latest run . --config '{"geminiApiKey":"..."}'

# Example: Run with Azure keys
npx -y @smithery/cli@latest run . --config '{"azureApiKey":"...", "azureApiBase":"https://your-azure.openai.azure.com/", "azureApiVersion":"2024-02-01", "azureDeploymentId":"your-deployment"}'

The smithery.yaml file defines the necessary configuration schema. Refer to it for all available options.