This Zotero plugin enables users to interact with their PDF research papers using various AI language models, streamlining the research and reading process. It targets scholars, researchers, and anyone managing large volumes of PDF documents, offering efficient insight extraction and question-answering capabilities.

How It Works

The plugin integrates with Zotero, allowing users to select multiple PDFs or entire collections. It leverages a Retrieval-Augmented Generation (RAG) approach, with local embeddings, vector databases, and reranking modules. This design ensures 100% privacy and offline functionality, as all processing occurs locally.

Quick Start & Requirements

• Installation: Download papersgpt.xpi and install via Zotero's Add-ons menu (Tools > Add-ons > Gear Icon > Install Add-on From File).
• Startup: Press Command + Enter (macOS) or Ctrl + Enter (Windows) within Zotero.
• Dependencies: For local LLMs on macOS, a minimum of 8GB RAM is recommended (12GB for Gemma 3). Windows users may need to deploy local LLMs via Ollama or llama.cpp and connect via a custom API.
• Documentation: GitHub Repository

Highlighted Details

• Supports a wide array of commercial LLMs via OpenRouter (GPT-4.1, Gemini 2.5 Pro, Claude Sonnet 3.7, etc.) and numerous SOTA open-source models locally on macOS (Qwen 3, Gemma 3, DeepSeek-R1-Distill-Llama, Llama 3.2, Mistral).
• Offers built-in prompts for common research tasks like literature reviews and theoretical frameworks.
• All RAG components (embeddings, vector database, reranking) run locally for enhanced privacy and offline use.
• Seamless integration within the Zotero interface for managing and querying documents.

Maintenance & Community

Contributions are welcome via standard GitHub pull requests and issue reporting.

Licensing & Compatibility

The README does not explicitly state the license. Compatibility for commercial use or closed-source linking is not specified.

Limitations & Caveats

The plugin is distributed as an .xpi file, implying it's a Zotero add-on; however, the specific license is not mentioned, which could impact commercial use. Windows users require additional setup for local LLM integration.