Zotero MCP is a Model Context Protocol server that connects your Zotero research library with Claude, ChatGPT, and other AI assistants to enable discussion, summarization, and analysis of your papers.
The tool solves the problem of isolated research libraries by bridging Zotero with AI assistants through the Model Context Protocol. It works by exposing your Zotero collection—whether stored locally or accessed via the web API—as a set of tools that AI assistants can call. This allows you to search papers by title, author, or semantic similarity, retrieve full metadata and content, extract PDF annotations with page numbers, and perform write operations like adding papers by DOI or URL with automatic metadata fetching and open-access PDF discovery.
Researchers managing substantial paper collections in Zotero should consider this tool if they want to discuss their library contents with AI assistants without manually copying citations or abstracts. The project suits workflows where you need semantic search across your library, want to extract and analyze annotations programmatically, or need to add papers to your collection through AI-assisted discovery. The tool supports multiple embedding models for semantic search, including a free local default option as well as OpenAI, Gemini, and Ollama, giving you flexibility in how you run similarity searches. Optional extras provide Scite citation intelligence to identify retracted papers and see citation tallies, and PDF table-of-contents extraction. A hybrid mode allows local-only users to perform reads locally while using the web API for write operations.
The project maintains active engagement with its user community through a dedicated Discord server. Development shows responsiveness to feature requests and bug reports, with the maintainer regularly updating the tool to support new AI assistant platforms and expanding integration capabilities. The codebase is distributed via PyPI for straightforward installation and updates.
