llm-for-zotero: what it is, what problem it solves & why it's gaining traction

What it solves

llm-for-zotero integrates Large Language Models (LLMs) directly into the Zotero PDF reader, eliminating the need to switch between your research library and an AI assistant. It allows researchers to summarize papers, ask questions about specific text or figures, and organize their library using AI agents without leaving the Zotero interface.

How it works

The plugin connects Zotero to various LLM backends, including standard API providers, local models, ChatGPT WebChat, Codex App Server, and Claude Code. It uses retrieval-augmented generation (RAG) to provide grounded answers with citations that link back to the source passage in the PDF.

For advanced workflows, it features an Agent Mode that can interact with the Zotero API to search the library, update metadata, apply tags, and manage collections. It also supports MinerU PDF parsing for high-fidelity extraction of tables and equations. Users can save AI-generated notes to Zotero's internal notes or export them as Markdown files to external tools like Obsidian or Logseq.

Who it’s for

Academic researchers, students, and anyone managing a large library of scholarly papers in Zotero who wants to automate literature review, paper analysis, and note-taking.

Highlights

• Grounded Paper Chat: Ask questions about the current PDF with citations that jump directly to the source text.
• Agent Mode: An AI agent capable of reading and writing to your Zotero library (tagging, metadata updates, importing references).
• Multi-Paper Comparison: Compare multiple open papers using slash commands to add them as context.
• External Note Integration: Save research notes directly to local Markdown folders (e.g., Obsidian, Logseq).
• Flexible Backends: Supports a wide range of providers, from API keys and local models to WebChat and Claude Code.
• Customizable Skills: Pre-defined or user-created guidance files to shape how the agent handles specific research tasks.