Zotero MCP Server

An MCP server that connects AI assistants to your Zotero library — including full-text PDF extraction. Search papers, browse collections, pull citations, and read the actual content of your PDFs, all from within Claude Code or any MCP-compatible client.

What makes this different

Most Zotero integrations only give you metadata. This one downloads your PDFs from Zotero cloud and extracts the full text using pymupdf, so your AI assistant can actually read your papers.

Setup

1. Get your Zotero credentials

API Key: Go to https://www.zotero.org/settings/keys/new and create a key with read access
User ID: Shown at https://www.zotero.org/settings/keys — the number at the top

2. Install dependencies

pip3 install httpx "mcp[cli]" pydantic pymupdf

3. Add to Claude Code

claude mcp add zotero-mcp \
  --command python3 \
  --args "/path/to/zotero-mcp/server.py" \
  --env ZOTERO_API_KEY=your_api_key_here \
  --env ZOTERO_USER_ID=your_user_id_here

Or add manually to ~/.claude.json:

{
  "mcpServers": {
    "zotero-mcp": {
      "command": "python3",
      "args": ["/path/to/zotero-mcp/server.py"],
      "env": {
        "ZOTERO_API_KEY": "your_api_key_here",
        "ZOTERO_USER_ID": "your_user_id_here"
      }
    }
  }
}

4. Restart Claude Code and test

Search my Zotero library for PCSK9

Tools

| Tool | What it does | |------|-------------| | zotero_search_library | Full-text search across titles, authors, abstracts, tags | | zotero_get_item | Full metadata for a specific item by key | | zotero_get_fulltext | Download PDF and extract full text (with optional page ranges) | | zotero_list_attachments | List all attachments (PDFs, snapshots) for an item | | zotero_list_collections | List all your collections (folders) | | zotero_get_collection_items | Browse papers in a collection | | zotero_search_by_tag | Find papers with a specific tag | | zotero_get_recent_items | Most recently added papers | | zotero_list_tags | All tags with item counts | | zotero_get_citation | Formatted citation (Vancouver, APA, AMA, NEJM, Chicago) |

Example prompts

Search my Zotero for PCSK9 heart failure papers
Get the full text of item AB12CD34
Read pages 1-5 of that paper
What papers do I have on statin intolerance?
List my Zotero collections
Show me papers tagged 'key paper'
Cite item AB12CD34 in Vancouver style
What did I add to Zotero recently?

Notes on full-text extraction

PDFs must be synced to Zotero cloud (not just linked locally). If you see "linked file" errors, right-click the attachment in Zotero desktop and select "Convert Linked File to Stored File", then sync.
Large papers can be read in chunks using page_start and page_end parameters.
Text extraction uses pymupdf which handles multi-column academic layouts well.

Requirements

Python 3.10+
A Zotero account with API access enabled
Zotero cloud storage (free tier: 300MB, or paid plans for more)

License

MIT

zotero-mcp