OpenViSTAI · Agents · Blockchain · Crypto
Submit a project
Menu
MCP & Tooling

Zotero MCP

Zotero MCP is a Python-based Model Context Protocol tool that connects Zotero research libraries with AI assistants. It offers an MCP server, command-line tools with an agent skill, and a Zotero chat plugin for searching, reading, citing and annotating papers.

Explore the practical guide ↓

Independent OpenViST project profile, based on linked sources.

Zotero MCP — OpenViST project profile in MCP & Tooling.
OpenViST project illustration. See the linked sources for product details.
Understand the project

Practical overview

When it fits

Use it when your research collection is in Zotero and you want an AI assistant to find, read or annotate items there. Choose the MCP server for clients that speak MCP, the CLI and agent skill for shell-enabled agents, or the Zotero plugin to work beside an open paper.

Who it is for

  • Researchers who manage papers in Zotero and want to search or discuss them with an AI assistant.
  • People using MCP-capable chat clients or shell-enabled agents with a Zotero library.
  • Zotero readers who want an assistant panel alongside an open paper.

Documented uses

  • Search a library by title, author, tag, collection, full text or meaning.
  • Read paper metadata, BibTeX, full text and PDF page ranges; retrieve page images when text extraction garbles figures, tables or math.
  • Discuss papers, summarize findings and analyze citations with an assistant.
  • Add papers by DOI, URL, ISBN, BibTeX or file, manage tags and collections, and merge duplicates.
  • Create highlights, area-box annotations and notes, or extract existing PDF annotations.
Before you choose

Requirements & limits

The documented quick start requires Python 3.10+ and Zotero 7+ with its local API enabled. Writes require separate authorization: local authorization on Zotero 10+, or a web API key and library ID on older versions. The repository lists an MIT license.

  • Three connection routes serve different working contexts: MCP clients, shell-enabled agents, and an in-Zotero chat panel.
  • Semantic search can use local, OpenAI, Gemini or Ollama embeddings.
  • Local mode reads from zotero.sqlite; writes use a running Zotero 10+ installation or the web API.
  • Scite citation tallies and retraction alerts are optional extras.
  • The README describes page-linked responses in the Zotero Agent panel, with citations that can open the cited page.

Check current requirements in the official documentation. Source-based guidance does not mean OpenViST has run or independently tested this project.

Source-based practical guide

Install Zotero MCP and enable local access

Prepare the package and Zotero local API for connecting an assistant.

Before you start

  • Python 3.10 or later
  • Zotero 7 or later

Steps

  1. Install the package with the command below or the command below.

    Documented command
    uv tool install zotero-mcp-server
    Documented command
    pip install zotero-mcp-server

  2. In Zotero, open Settings → Advanced and enable “Allow other applications on this computer to communicate with Zotero.”

  3. For Claude Desktop, run zotero-mcp setup to configure the connection.

Expected outcome

The package is installed, Zotero's local API is enabled, and Claude Desktop has been configured to connect to the MCP server.

Guide limitations

  • This guide covers Claude Desktop's documented setup path; other clients have separate instructions.
  • Writing to the library requires an additional authorization step. On Zotero 10+, run zotero-mcp authorize-local and choose Always Allow; older Zotero versions use a web API key and library ID.
Documentation 1 ↗
Evidence and transparency

Sources & review

Official repository ↗
Source record details
Source availability
Active
Sources checked
9 October 2026
Original website title
54yyyu/zotero-mcp · Official homepage title · retained in its source language
How we review sourcesSuggest a correction