# PapersGPT ## Industrial-Scale Multi-Document Intelligence for Zotero. 100% Local. Zero Hallucinations. **PapersGPT** is a professional-grade Zotero extension engineered in Native C++ to transform thousands of isolated PDFs into a single, high-precision knowledge engine. While standard tools struggle with single-file chat or slow, "fuzzy" embeddings, PapersGPT delivers instant, cross-document synthesis and global library search with absolute data sovereignty. --- ## ⚡ Extreme Performance: Built for "Big Data" Researchers Stop waiting for embedding progress bars. Most AI tools use Python/JS scripts that lag or crash as your library grows. **PapersGPT’s C++ core provides:** - **Near-Instant Indexing**: Process 1,000+ documents in minutes, not hours. - **Hardware Efficiency**: Millisecond-level retrieval with minimal RAM usage, leaving your system resources free for other tasks. - **Zero-Cloud Dependency**: High-speed structural indexing occurs 100% on your machine. No external APIs, no cloud bottlenecks, even works in Airplane Mode. --- ## 🎯 Superior Multi-Doc Intelligence & Accuracy PapersGPT moves beyond the "fuzzy" matching of standard semantic search by embeddings. Our engine understands document structure, ensuring that cross-paper comparisons are logically sound and factually accurate. - **Deep Cross-Document Comparison** *“Compare the sample sizes and p-values of all 50 clinical trials in my 'Cardiology' folder.”* AI synthesizes data from multiple files into structured tables or reports without losing the nuances of individual studies. - **Global Library Synthesis** Ask questions across your entire Zotero library. Identify trends, evolution of terms, or conflicting evidence across 10,000+ documents simultaneously. - **Verifiable Evidence (Click-to-Jump)** Accuracy is useless without proof. Every synthesis includes interactive citations. Every AI response includes precise citations; clicking on them takes you directly to the specific paper being cited. - **Ensure the absolute 100% privacy and security of your local sensitive Zotero data** - **Support 100% offline local LLMs** PapersGPT currently supports three different ways to connect to local large language models (LLMs): 1. **Integrated SOTA open-source LLMs** – such as Gemma 4 12B and Qwen3.5 4B. These models can be downloaded and run with a single click, eliminating the need for complex installation procedures; simply selecting a model from the "Local LLM" option triggers an automatic download from Hugging Face and launches the model. 2. **Custom OpenAI-compatible endpoint** – connect to a flexible OpenAI-compatible privately deployed LLM endpoint by configuring a customized LLM API within PapersGPT. 3. **Ollama integration** – supports connections to local Ollama instances; any LLM deployed and running locally via Ollama can be connected and used simply by configuring it within PapersGPT. - **Parsing, Indexing and Searching** documents in Zotero are performed locally, there is no reliance on the clouds, and these operations can be run offline. - **Local LLMs and local document processing** ensure **"zero-byte" data leakage** from your device. --- ## ✨ Core Unique Features - **AutoPilot Automation**: The world's first "Autonomous Researcher" for Zotero. Define a research goal and let AI batch-process 1,000+ papers overnight, saving insights directly into Zotero Notes while you sleep. - **SOTA Model Cowork**: Seamlessly switch between Claude 4.7, GPT-5.5, and DeepSeek, or run Free Local LLMs (via Ollama) for a total air-gapped experience. - **Zotero MCP Server**: Integrate your local knowledge base as a primary "skill" for agentic platforms like Claude Code, Cursor, or Windsurf. --- ## 📖 Quickstart ### Step 1: Download and Install 1. First, download the latest [PapersGPT](https://github.com/papersgpt/papersgpt-for-zotero/releases/download/papersgpt-v0.6.0/papersgpt-v0.6.0.xpi). 2. Then, install the downloaded Zotero plugin file. For more details, please see [here](https://www.papersgpt.com/quickstart). --- ### Step 2: Start Chatting with a PDF, Multiple PDFs or your entire Zotero Library #### 1. Launch PapersGPT - **Chat with a Single PDF** - Open a PDF from your Zotero library. - When you need AI assistance, click papersgpt-logo on the top menu of the PDF viewer or use a keyboard shortcut: - **macOS:** `Command + Enter` - **Windows:** `Ctrl + Enter` - **Chat with Multiple PDFs** - Select multiple files or a collection in the main Zotero window. Hold `Ctrl` while clicking files on Windows. Hold `Command` while clicking files on Mac. - **Chat with the entire Zotero Library** - Simply check "Search entire library" in PapersGPT, and it will search the entire document library for relevant documents and synthesize answers based on the questions. #### 2. Select a LLM model and configure the API KEY of the model For more detailed information please see [here](https://www.papersgpt.com/models). #### 3. Ask Questions - Use the built-in prompts for common tasks like: Summary, Background, Generating a literature review, Theoretical frameworks, Future directions. - You can also directly type any question or custom prompt to start the conversation. --- ### Step 3: Manage Your Findings and close the chat - After chatting, you can easily save the key insights and answers you've gathered from the conversation. - When you're finished, click the red cross (X) close button to exit the PapersGPT window. --- ## 🔌 MCP & SKILL Support **Support MCP** which makes you can access your Zotero papers in Claude, ChatGPT desktop app, please see [here](https://github.com/docsagent/docsagent) in detail. **Professional [SKILL](papersgpt-for-zotero/SKILL.md)** provides various agents for connecting to your local Zotero library. If you use OpenClaw, you can see the papersgpt-for-zotero skill [here](https://clawhub.ai/papersgpt/papersgpt-for-zotero). You are welcome to submit any issues in this github repo or clawhub. If you use the App of OpenAI, Claude, Cursor, Claude Code, CodeX or Gemini CLI, you can directly copy and load the [SKILL](papersgpt-for-zotero/SKILL.md). --- ## 🏢 Professional Compliance PapersGPT is designed for professionals in **Research, Legal and Finance** who handle high-stakes, sensitive data: - **Research**: Cross-reference thousands of papers while maintaining 100% data sovereignty. - **Finance**: Synthesize decade-long 10-K filings without cloud risk. - **Legal**: Perform high-precision discovery across thousands of case files. --- ## 👥 Join 10,000+ Researchers **Unlock Professional Capabilities** at [papersgpt.com](https://www.papersgpt.com) → > Don't just search your library. Understand it. PapersGPT provides the industrial-scale intelligence your research library demands.