SafeRag Documentation

Welcome to the SafeRag documentation. Learn how to set up and use your private AI assistant that runs entirely on your Mac.

Quick Links

💻 Installation Guide 💬 Chat Interface 📃 Document RAG 🤖 AI Models ⚙ Settings 🛠 Troubleshooting

What is SafeRag?

SafeRag is a private AI chat application for macOS that runs entirely on your computer. Unlike cloud-based AI services, SafeRag processes all conversations locally using Ollama, ensuring your data never leaves your Mac.

🔒

100% Private

All AI processing happens on your Mac. No data is sent to external servers.

📄

Document Intelligence

Upload documents and chat with them using RAG (Retrieval-Augmented Generation).

✅

Compliance Ready

Built-in GDPR and HIPAA compliance features with audit logging.

👥

Multi-User Support

Support for multiple users with role-based access control.

Key Features

Chat Modes

SafeRag offers multiple chat modes to suit different needs:

  • General Mode - Standard AI conversations for everyday tasks
  • RAG Mode - Chat with your uploaded documents for context-aware responses
  • Spark Mode - Creative mode with enhanced temperature for brainstorming

Document Support

Upload and chat with various document types:

  • PDF documents
  • Plain text files (.txt)
  • Markdown files (.md)
  • Images (PNG, JPEG) for visual context

Pro Features Pro

SafeRag Pro unlocks additional capabilities:

  • Unlimited Documents - No limit on RAG document uploads (Free: 5 documents)
  • Watch Folders - Automatically index documents from specified folders
  • Multi-User - Support up to 5 users with admin controls
  • Audit Logging - Complete activity logs for compliance
  • Data Export - Export all user data (GDPR Article 15)
💡 Free Trial
SafeRag includes a 24-hour free trial of all Pro features. No credit card required.

System Requirements

Requirement Minimum Recommended
macOS Version macOS 14.0 (Sonoma) macOS 14.0 or later
Processor Intel or Apple Silicon Apple Silicon (M1/M2/M3/M4)
RAM 8 GB 16 GB or more
Storage 5 GB free 20 GB+ for multiple models
✅ Apple Silicon Recommended
For the best performance, we recommend Apple Silicon Macs (M1, M2, M3, or M4). These processors provide significantly faster AI inference compared to Intel-based Macs.

Getting Started

Ready to begin? Follow these steps:

Download and Install

Download SafeRag from our website and install it on your Mac. See the Installation Guide for detailed instructions.

Complete Onboarding

The app will guide you through setup, including Ollama installation and downloading your first AI model. See Onboarding.

Start Chatting

Create your first chat session and start having private AI conversations. Learn more in Chat Interface.