Semantic Scholar API MCP server
LocalAbout Semantic Scholar API MCP server
An MCP server for searching papers through the Semantic Scholar API, built with the mcp Python SDK. It suits researchers who want literature search inside Claude or Cline. Setting an API key raises the rate limits.
Key features
- 01Paper search via Semantic Scholar
- 02Optional API key for higher rate limits
- 03Run with mcp dev or mcp install
- 04Configuration for Claude and Cline
Use cases
- 01Assisting literature review within an AI assistant
- 02Serving as a paper search tool for a pre-research agent
MCP support
This is an MCP server that provides Semantic Scholar paper search to MCP clients such as Claude and Cline.
Configuration
Local server config is the same for every client. Copy it into your client’s MCP config file.
We haven’t collected a package or remote endpoint for this server. Follow the instructions in the project README.
Tools
Local servers expose tools only once running on your machine, and the maintainer has not provided a tool list.
Details
From the project READMESemantic Scholar API MCP server
Made with mcp-python-sdk
Important
if you are still using FastMCP version of this mcp server, please consider pull this repo again and update to newer versions as FastMCP is already deprecated.
Usage
Requirements: pip install -r requirements.txt
Run mcp dev path/to/semantic-scholar-plugin.py to initialize the server.
Run mcp install path/to/semantic-scholar-plugin.py to install to claude or add following to claude/cline config:
"semantic-scholar": {
"command": "uv",
"args": [
"run",
"--with",
"mcp",
"mcp",
"run",
"/path/to/semantic-scholar-plugin.py"
]
}
Note
Currently using
uvwithmcpseems to break certain Linux/macOS version of Claude-desktop, you might need to set as:"semantic-scholar": { "command": "/path/to/mcp", "args": [ "run", "/path/to/semantic-scholar-plugin.py" ] }instead, with
/path/to/mcpgot from runningwhich mcpin terminal
API Key
To use the Semantic Scholar API with higher rate limits, you can set your API key as an environment variable:
export SEMANTIC_SCHOLAR_API_KEY="your_api_key"
or set by adding an env key in mcp settings by:
"semantic-scholar": {
"command": ...,
"args": ...,
"env": {
"SEMANTIC_SCHOLAR_API_KEY": "your_api_key"
}
}
You can get an API key by filling out the form at: https://www.semanticscholar.org/product/api
Suggested Agent System prompt
See: benhaotang/my_agent_system_prompt, the AI pre-research agent that can make full use of this mcp server.
Known issues
- If you see things like
INFO Processing request of type __init__.py:431 ListToolsRequestin cline, you can ignore them as this will not affect it from working, this is because cline parse tool list together with console debug infos, and current python-sdk cannot disable console messages. This will not affect any function calling part other than seeing this warning.
Comments
New to MCP?
From what the protocol is to installing your first server, these answers cover most first questions.
An open protocol that defines how AI applications connect to external tools and data in a uniform way. Clients talk to the model; MCP servers provide tools, resources and prompt templates. They exchange standard messages, so one server can be reused by many clients.