SciSpace MCP Server

Search and explore 280M+ academic research papers from directly inside your AI assistant. The SciSpace MCP server brings peer-reviewed literature, abstracts, methodologies, results, and AI-generated paper insights into Claude and ChatGPT.

The SciSpace MCP server is currently available only through the official Claude and ChatGPT connector directories.

What it does

SciSpace MCP exposes the SciSpace research engine over the Model Context Protocol so your AI assistant can:

Run vector-based semantic search over hundreds of millions of academic papers.
Pull titles, authors, year, citations, PDF links, and full abstracts for each result.
Add structured analytical columns on demand — e.g. Summarized Abstract, Methods Used, Results, Limitations, Contributions, Research Gap, Future Research, Dataset, Independent / Dependent Variables, Problem Statement, Challenges, Applications, and more.
Render results in an interactive table widget (in clients that support MCP UI widgets) with the option to download the table.

Use it for literature reviews, background research, citation hunting, comparing methodologies across papers, or summarizing the state of a field.

Available tools

Tool	What it does
`search-papers`	Takes a natural-language research question and returns the most relevant papers with titles and abstracts.
`add-column`	Adds an additional analytical column (e.g. Methods, Limitations, Results) to a previously returned set of papers.

Setting it up

Claude (claude.ai and Claude Desktop)

Open Settings → Connectors.
Find SciSpace in the connector directory and click Connect.
Start a new chat and ask a research question — Claude will use the SciSpace tools automatically when relevant, or you can invoke them explicitly.