Chat with Your Research Library

Seamlessly connect your Zotero library with Claude, ChatGPT, and other AI assistants via the Model Context Protocol. Search, summarize, and analyze your research with AI.

Zotero Claude ChatGPT MCP
Get Started Explore Features
Works with your favorite AI clients
Claude Desktop ChatGPT Cherry Studio Cursor Chorus

Powerful Features

Everything you need to supercharge your research workflow with AI

AI-Powered Semantic Search

Go beyond keyword matching. Our vector-based semantic search understands the meaning behind your queries, finding conceptually related papers even when they use different terminology.

  • Multiple embedding models: Local (free), OpenAI, Gemini, HuggingFace
  • Intelligent similarity scoring
  • Auto-updating database with configurable sync
  • Full-text content indexing option
# Example semantic search queries
"Find papers similar to machine learning in neuroscience"
"Research about climate impacts on agriculture"
"Studies on social media and mental health"
"Papers conceptually similar to this abstract: [text]"
Smart Library Search

Find papers by title, author, content, or tags. Advanced multi-criteria search with AND/OR logic and tag filtering.

Full Content Access

Retrieve detailed metadata, full-text content, and attachments. Extract and analyze your research papers with AI.

Annotations & Notes

Extract PDF annotations, search notes, and create new research notes. Works with Zotero's native annotations.

Write Operations

Add papers by DOI or URL, manage collections, update metadata, batch tag operations, and find & merge duplicates.

BibTeX Export

Export citations in BibTeX format directly through your AI assistant. Perfect for academic writing workflows.

Smart Updates

Auto-detect installation method and preserve configurations. Install optional extras like [semantic], [pdf], or [all] for additional functionality.

Quick Start

Get up and running in under 2 minutes

Recommended: UV Installation

The fastest way to get started with Zotero MCP:

# Core install
uv tool install zotero-mcp-server

# Full install (semantic search + PDF extraction)
uv tool install "zotero-mcp-server[all]"

zotero-mcp setup

Requires uv package manager

Pip Installation

Classic Python installation:

# Core install
pip install zotero-mcp-server

# Full install (semantic search + PDF extraction)
pip install "zotero-mcp-server[all]"

zotero-mcp setup
Optional Extras

Install only the extras you need:

  • [semantic] - Semantic search with vector embeddings (ChromaDB, embedding models)
  • [pdf] - PDF text extraction and outline parsing
  • [all] - Everything above combined
Enable Semantic Search

Set up AI-powered semantic search:

# Configure semantic search
zotero-mcp setup --semantic-config-only

# Build the search database
zotero-mcp update-db --fulltext
Full Installation Guide

Setup & Usage

Configure for your favorite AI client and start chatting with your research

Claude Desktop
Auto-configure (recommended):
zotero-mcp setup
Manual configuration:

Add to your claude_desktop_config.json:

{
  "mcpServers": {
    "zotero": {
      "command": "zotero-mcp",
      "env": {
        "ZOTERO_LOCAL": "true"
      }
    }
  }
}
Cherry Studio

Go to Settings → MCP Servers → Edit MCP Configuration:

{
  "mcpServers": {
    "zotero": {
      "name": "zotero",
      "type": "stdio",
      "isActive": true,
      "command": "zotero-mcp",
      "args": [],
      "env": {
        "ZOTERO_LOCAL": "true"
      }
    }
  }
}
Example Prompts
Basic Search
  • "Search my library for papers on machine learning"
  • "Find recent articles about climate change"
  • "Show papers tagged #important"
  • "Export BibTeX for my quantum computing papers"
Semantic Search
  • "Find papers conceptually similar to deep learning in vision"
  • "Research about AI applications in healthcare"
  • "Papers discussing topics like this abstract: [paste text]"
Analysis & Summarization
  • "Summarize the key findings from my paper on neural networks"
  • "Extract all annotations from my quantum computing paper"
  • "Compare the methodologies across my ML papers"
Organization
  • "Tag all my deep learning papers with 'AI' and 'neural-networks'"
  • "Create a summary note for this paper"
  • "Search my notes for mentions of 'reinforcement learning'"
Write & Manage
  • "Add this paper by DOI: 10.1234/example"
  • "Create a collection called 'Literature Review'"
  • "Find and merge duplicate papers in my library"
  • "Tag all papers in my 'AI' collection with 'reviewed'"

Advanced Configuration

Fine-tune Zotero MCP for your workflow

Web API Access

Access your library remotely via Zotero's Web API:

zotero-mcp setup --no-local \
  --api-key YOUR_API_KEY \
  --library-id YOUR_LIBRARY_ID

Get your API key at zotero.org/settings/keys

Environment Variables
Zotero Connection
  • ZOTERO_LOCAL=true - Use local API
  • ZOTERO_API_KEY - Web API key
  • ZOTERO_LIBRARY_ID - Library ID
  • ZOTERO_LIBRARY_TYPE - user or group
Semantic Search
  • ZOTERO_EMBEDDING_MODEL - default, openai, gemini, huggingface
  • OPENAI_API_KEY - For OpenAI embeddings
  • GEMINI_API_KEY - For Gemini embeddings
CLI Commands
Setup & Server
zotero-mcp setup              # Auto-configure
zotero-mcp setup-info         # Show config info
zotero-mcp serve              # Run server
zotero-mcp serve --transport streamable-http
Updates & Database
zotero-mcp update             # Update to latest
zotero-mcp update --check-only
zotero-mcp update-db          # Update search DB
zotero-mcp update-db --fulltext
zotero-mcp db-status          # Check DB status

Available Tools

Comprehensive API for interacting with your Zotero library

Semantic Search
  • zotero_semantic_search
  • zotero_update_search_database
  • zotero_get_search_database_status
Search Tools
  • zotero_search_items
  • zotero_search_by_tag
  • zotero_advanced_search
  • zotero_get_collections
  • zotero_get_collection_items
  • zotero_get_tags
  • zotero_get_recent
  • zotero_search_by_citation_key
Content Tools
  • zotero_get_item_metadata
  • zotero_get_item_fulltext
  • zotero_get_item_children
  • zotero_get_pdf_outline
Notes & Annotations
  • zotero_get_annotations
  • zotero_get_notes
  • zotero_search_notes
  • zotero_create_note
  • zotero_create_annotation
Write & Management
  • zotero_add_by_doi
  • zotero_add_by_url
  • zotero_add_from_file
  • zotero_create_collection
  • zotero_update_item
  • zotero_find_duplicates
  • zotero_merge_duplicates
  • zotero_batch_update_tags
  • zotero_search_by_citation_key
Full API Documentation

FAQ & Troubleshooting

Common issues and their solutions

  1. Ensure Zotero is running
  2. Enable the local API: Go to Edit > Preferences > Advanced and toggle on "Allow other applications on this computer to communicate with Zotero"
  3. If using Web API, verify your API key and library ID

  1. Run zotero-mcp setup to configure your environment
  2. Initialize the database: zotero-mcp update-db
  3. Check status: zotero-mcp db-status
  4. For better results, use --fulltext flag when building the database
  5. If you have embedding conflicts, use --force-rebuild

  • Requires Zotero 7+ for local full-text access
  • Not all items may have full-text content available
  • Web API has limited attachment access

For optimal annotation extraction, install the Better BibTeX plugin for Zotero.

As a fallback, try using the use_pdf_extraction: true parameter with the zotero_get_annotations tool.

Problems from changing install methods or search options can often be resolved with:

zotero-mcp update-db --force-rebuild

For update issues, try:

zotero-mcp update --force

If you changed embedding models and are seeing 404 or NOT_FOUND errors, the existing database was built with a different model. Rebuild it:

zotero-mcp update-db --force-rebuild

If that doesn't resolve the issue, delete the database directory and rebuild from scratch:

rm -rf ~/.config/zotero-mcp/chroma_db/
zotero-mcp update-db --fulltext