home / mcp / semantic scholar mcp server

Semantic Scholar MCP Server

Provides access to the Semantic Scholar API with higher rate limits via MCP for plug-in usage in Claude.

Installation

Add the following to your MCP client configuration file.

Configuration

{
  "mcpServers": {
    "benhaotang-mcp-semantic-scholar-server": {
      "command": "uv",
      "args": [
        "run",
        "--with",
        "mcp",
        "mcp",
        "run",
        "/path/to/semantic-scholar-plugin.py"
      ],
      "env": {
        "SEMANTIC_SCHOLAR_API_KEY": "YOUR_API_KEY"
      }
    }
  }
}

This MCP server provides a Semantic Scholar API integration that you can run and install to Claude via MCP. It enables you to access the Semantic Scholar data with higher rate limits and manage the plugin lifecycle from your MCP client.

How to use

To use this MCP server, first initialize the server for development and then install it to your target environment (such as Claude). You can start a development session to test the plugin, and once you’re satisfied, install it so your MCP client can run the semantic scholar plugin as part of your workflow.

Usage patterns include starting a development server for testing, then installing the plugin so it can be run by your MCP client. If you need higher API rate limits, provide your API key as an environment variable (see API Key section) to ensure the plugin talks to Semantic Scholar with elevated quotas.

How to install

Prerequisites: make sure you have Python installed and the required dependencies installed.

Step 1: Install dependencies from the requirements file.

Step 2: Initialize the server for development.

Step 3: Install the MCP server to your environment (Claude or your MCP client) using the provided paths.

Step 4: If you prefer using a direct MCP runtime, you can configure Claude with the following standard MCP config entries.

Complete steps are shown in the code blocks below for the two MCP run configurations.

Additional sections

API Key support lets you increase rate limits by supplying an API key. You can provide the key as an environment variable or via the MCP server settings.

Environment variable option: set SEMANTIC_SCHOLAR_API_KEY in your shell or in the MCP settings under env for the semantic_scholar server.

Known issues: some console logs may appear during tool listing, but they do not affect functionality.

Config snippets

{
  "semantic_scholar": {
    "command": "uv",
    "args": [
      "run",
      "--with",
      "mcp",
      "mcp",
      "run",
      "/path/to/semantic-scholar-plugin.py"
    ],
    "env": {
      "SEMANTIC_SCHOLAR_API_KEY": "YOUR_API_KEY"
    }
  }
}

{
  "semantic_scholar": {
    "command": "/path/to/mcp",
    "args": [
      "run",
      "/path/to/semantic-scholar-plugin.py"
    ],
    "env": {
      "SEMANTIC_SCHOLAR_API_KEY": "YOUR_API_KEY"
    }
  }
}