arxiv_paper_mcp

JminJ/arxiv_paper_mcp

3.2

If you are the rightful owner of arxiv_paper_mcp and would like to certify it and/or have it hosted online, please leave a comment on the right or send an email to henry@mcphub.com.

The MCP Server is designed to facilitate the search of academic papers from the Arxiv API using Google's LLM technology.

Tools
1
Resources
0
Prompts
0

Arxiv Paper MCP Server

MCP(Model Context Protocol) server can search papersfrom arxiv api.

1. Pre-Required

  1. Google llm api key : This mcp server use google llm(gemini series). So generate google llm api key first.
  2. Select using model : This mcp server initialize llm model by langchain's init_chat_model function. Follow this format "{model_provider}:{model_name}"; Example value "google_genai:gemini-2.0-flash".

2. Installation

Cursor

uv sync

source .venv/bin/activate

# generate mcp.json setting
python -m src.setup.cursor \
    -e GOOGLE_API_KEY=<YOUR GOOGLE LLM API KEY> \
    -e USING_MODEL_INFO=<YOUR USING MODEL INFO> \
    -e MODEL_TEMPERATURE=0.0 \
    -e MODEL_SEED=256

3. Support Tool

a. search_papers_based_user_query

Search papers from Arxiv API that user want to search. Example

Q: I want to search recent papers in NLP domain.

b. paper_download_tool

Download paper by user given arxiv id. Example

Q: Download 1706.03762 paper.

c. return_section_names_tool

Return section names of target paper. Example

Q: showing section of 1706.03762 paper to me.

d. description_paper_content

Descrive paper content and section from user question. Example

Q: Descrive "Model Architecture" section to me.

4. License

This project is open source software licensed as MIT.