DEV.co
RAG Frameworks · yilewang

llm-for-zotero

llm-for-zotero is a TypeScript-based Zotero plugin that integrates LLM capabilities directly into the PDF reader, enabling researchers to chat with papers, summarize content, compare sources, and organize findings without leaving Zotero. It supports multiple LLM backends (OpenAI-compatible APIs, local models, WebChat, Codex, Claude Code) and features RAG-based grounded answers with citation navigation.

Source: GitHub — github.com/yilewang/llm-for-zotero
2.2k
GitHub stars
113
Forks
TypeScript
Primary language
AGPL-3.0
License (OSI-approved)

Key facts

Objective fields from the source. Values we can't verify are shown as “Unknown” rather than guessed.

FieldValue
Repositoryyilewang/llm-for-zotero
Owneryilewang
Primary languageTypeScript
LicenseAGPL-3.0 — OSI-approved
Stars2.2k
Forks113
Open issues97
Latest releasev3.8.25 (2026-07-07)
Last updated2026-07-07
Sourcehttps://github.com/yilewang/llm-for-zotero

What llm-for-zotero is

A Zotero 7–9 plugin built in TypeScript using the Zotero plugin template, implementing RAG and agent-mode workflows with support for multiple LLM provider protocols (OpenAI chat compat, Anthropic messages, Gemini native). Includes PDF parsing (native and MinerU), multi-document context via slash commands, file uploads (PDF, DOCX, PPTX, TXT, Markdown), and local Markdown note export (Obsidian/Logseq compatible).

Quickstart

Get the llm-for-zotero source

Clone the repository and explore it locally.

terminalbash
git clone https://github.com/yilewang/llm-for-zotero.gitcd llm-for-zotero# follow the project's README for install & configuration

Need it deployed, integrated, or customized instead? DEV.co ships production installs.

Best use cases

Literature Review & Paper Summarization

Researchers can instantly summarize papers, compare methodology across multiple open PDFs, and extract key findings with citation-grounded answers that link back to source passages.

Research Note-Taking & Knowledge Management

Save AI-generated summaries and conversation history directly to Zotero notes or external Markdown folders (Obsidian, Logseq), enabling integrated research workflows without context switching.

Library-Wide Research Automation

Agent Mode with custom skills enables automated tagging, metadata editing, bulk import, and library-wide search and organization tasks across the entire Zotero collection.

Implementation considerations

  • Requires Zotero 7+ installed; plugin is installed via .xpi file drop-in (no build step for end users). Administrator must configure LLM provider (API key, base URL, model name) in preferences and test connectivity before use.
  • LLM backend selection is critical: API-based providers (OpenAI, Anthropic, Gemini) incur per-request costs and require internet; local models or WebChat avoid API fees but demand local infrastructure or browser automation.
  • Data privacy depends entirely on chosen backend—PDF content and chat history are sent to the configured LLM provider. Organizations handling sensitive research must review provider terms and consider local/on-premise models.
  • Multi-model configuration is supported (e.g., multimodal for figures, text-only for summaries); testing different model-task pairings may be needed to optimize quality and cost.
  • Agent Mode is beta and extends library-wide operations (read, search, tag, edit); thorough testing recommended before enabling in production workflows to avoid unintended bulk changes.

When to avoid it — and what to weigh

  • AGPL-3.0 Incompatibility — If your organization cannot comply with AGPL-3.0 (copyleft; derivative works must be released under AGPL), avoid this plugin. Proprietary or closed-source integrations require legal review.
  • No Commercial SLA Required — This is community-maintained open source with no guaranteed uptime, support response time, or commercial indemnification. Enterprise support contracts are not clearly offered.
  • Zotero Versions Before 7 — Plugin supports only Zotero 7, 8, and 9. Legacy Zotero installations or deeply customized Zotero forks may face compatibility issues.
  • Offline-Only Environments — Most backends (OpenAI API, Claude, Gemini) require internet access. Local model support exists but requires separate infrastructure (e.g., local OpenAI-compatible server or MinerU-API).

License & commercial use

Licensed under AGPL-3.0 (GNU Affero General Public License v3.0). This is a copyleft license requiring that any modified versions or derivative works must also be distributed under AGPL-3.0, and networked use (e.g., plugin running on a shared server) may trigger disclosure obligations. Requires legal review for commercial or proprietary use.

AGPL-3.0 is NOT a permissive OSI license suitable for closed-source or proprietary commercial use without legal review. If you plan to resell, embed, or modify this plugin for commercial purposes, consult legal counsel before proceeding. Commercial support, warranty, and indemnification are not clearly offered; community support is primary.

DEV.co evaluation signals

Editorial assessment — not user reviews. Directional, with an explicit confidence level.

SignalAssessment
MaintenanceActive
DocumentationStrong
License clarityClear
Deployment complexityLow
DEV.co fitGood
Assessment confidenceHigh
Security considerations

PDF content and chat history are sent to configured LLM backends; ensure compliance with data residency and privacy regulations (e.g., GDPR, HIPAA if applicable). No encryption at rest documented for local Zotero notes or chat history. Plugin runs inside Zotero process with full read access to library; review Zotero plugin sandboxing model. WebChat mode uses browser automation (potential for side-channel attacks). Requires review for handling of API keys and credential storage in Zotero preferences.

Alternatives to consider

Zotero native features + ChatGPT/Claude browser plugins

Simpler, avoids AGPL copyleft, but requires manual copy-paste of PDFs and lacks RAG grounding within Zotero. No citation navigation back to source.

Elicit, Scite, or ResearchRabbit

Commercial platforms with native LLM and literature-search integration, SLA support, and proprietary research APIs. No open-source copyleft restrictions, but higher cost and vendor lock-in.

Self-hosted RAG stack (LangChain + local LLM + Vector DB)

Full control, no API costs, no AGPL constraints, but requires significant engineering overhead and DevOps maintenance to integrate with Zotero.

Software development agency

Build on llm-for-zotero with DEV.co software developers

Download llm-for-zotero today and start chatting with your papers. Review the AGPL-3.0 license and commercial terms with your legal team if deploying in an enterprise or proprietary setting.

Talk to DEV.co

Related open-source tools

Surfaced by semantic similarity across the DEV.co open-source index.

Related on DEV.co

Explore the category and the services that help you build with it.

llm-for-zotero FAQ

Do I need an API key to use llm-for-zotero?
No. You can use WebChat (browser-based ChatGPT), Codex App Server (for ChatGPT Plus users), or a local OpenAI-compatible model. API-based providers (OpenAI, Claude, Gemini) do require keys and incur costs.
Can I use this plugin in a commercial research tool or SaaS product?
Not without legal review. AGPL-3.0 requires derivative works and networked applications to be open-source under the same license. Proprietary use is not permitted without a commercial license or relicensing agreement (which is not documented).
How does the plugin handle sensitive research data?
Data flows to the configured LLM backend (OpenAI, Anthropic, etc.). Local models or on-premise solutions keep data on-device. Review the backend's privacy policy and data retention terms. HIPAA, FERPA, or classified research may require dedicated compliance analysis.
Is Agent Mode production-ready?
Agent Mode is marked beta and enables library-wide operations (read, tagging, metadata editing). Test thoroughly before enabling in production; unintended bulk changes are possible. Skills are customizable but require manual configuration.

Custom software development services

Need help beyond evaluating llm-for-zotero? DEV.co is a software development agency offering software development services and web development for teams of every size. Our software developers and web developers build custom software, web applications, APIs, and rag frameworks integrations — and maintain them long-term.

Ready to Supercharge Your Research Workflow?

Download llm-for-zotero today and start chatting with your papers. Review the AGPL-3.0 license and commercial terms with your legal team if deploying in an enterprise or proprietary setting.