Q2 Edge Chat

Q2 Edge Chat is a privacy-focused chat application designed to run large language models directly on your iPhone. This app puts the power of generative AI models in your hands, allowing you to experiment with various models, ask questions, and explore prompts while keeping your data under your control. With Q2 Edge Chat, you can enjoy the benefits of advanced AI technology without compromising your privacy.

Benefits

Q2 Edge Chat offers several key advantages that set it apart from other chat applications:

• Privacy First: All conversations stay on your device, ensuring that your data is never sent to external servers. This makes Q2 Edge Chat an excellent choice for users who prioritize privacy and security.
• Offline Capable: Once models are downloaded, you can use the app without an internet connection, making it convenient for use in areas with limited connectivity.
• Model Management: The app integrates with HuggingFace, allowing you to browse and download a wide range of models. It also supports GGUF, optimized for quantized models, and offers a curated selection of high-quality small models.
• Modern Chat Experience: Q2 Edge Chat provides a smooth and responsive user interface with features like multiple conversations, dynamic text input, and persistent chat history. The app also supports dark and light modes, making it easy on the eyes.

Use Cases

Q2 Edge Chat is versatile and can be used in various scenarios:

• Personal Assistant: Use the app to ask questions, get information, and manage tasks without worrying about your data being collected or shared.
• Research: Explore different system prompts and experiment with various models to find the best fit for your needs.
• Education: Use the app to learn about different AI models and their capabilities, making it a valuable tool for students and educators.
• Creative Writing: Generate ideas, brainstorm, and get inspiration from AI models to enhance your creative writing process.

Technical Architecture

Q2 Edge Chat is built with a robust technical architecture that ensures smooth performance and easy model management. The app's core components include:

• LlamaEngine: A local inference engine for running GGUF models.
• ModelManager: Handles model downloading and file management.
• ChatManager: Manages conversations and model switching.
• ManifestStore: Persistent storage for downloaded models.

Getting Started

To get started with Q2 Edge Chat, you need an iPhone running iOS 18.0 or later or a Mac running macOS 13.0 or later. The app requires at least 4GB of available storage for models. You can clone the repository, open the project in Xcode, and build and run the app. Once installed, you can download your first model from the Browse Models screen.

Recommended Models

Q2 Edge Chat includes a curated selection of efficient models to help you get started:

• Phi-3.5 Mini (3.8B): Ideal for general conversation.
• Qwen2.5-1.5B (1.5B): Offers fast responses.
• SmolLM2-1.7B (1.7B): Great for code assistance.
• TinyLlama-1.1B (1.1B): An ultra-lightweight option for quick interactions.

Usage

Using Q2 Edge Chat is straightforward. Upon first launch, you'll see a welcome screen showing your model count and app status. You can then browse and download GGUF models, select a model, and start chatting. The app allows you to manage multiple conversations and export important chats for future reference.

Storage Requirements

Q2 Edge Chat has specific storage requirements to ensure optimal performance:

• App Size: ~50MB
• Model Storage: 1-4GB per model, depending on size and quantization.
• Chat History: ~1MB per 1000 messages.

Privacy & Security

Q2 Edge Chat is designed with privacy as its core principle. The app does not make network requests during conversations, and all processing happens on your device. There are no analytics or telemetry features, and your chat history is stored locally only. Models are cached locally after download, ensuring that your data remains secure.

Contributing

Q2 Edge Chat welcomes contributions from the community. You can fork the repository, create a feature branch, make your changes, test thoroughly on your device, and submit a pull request. This collaborative approach helps improve the app and make it more versatile for users.

License

Q2 Edge Chat is licensed under the Apache License, Version 2.0. This license allows for the free use, modification, and distribution of the software, making it accessible to a wide range of users and developers.

Support

For questions, issues, or feature requests, you can open an issue on GitHub. The Q2 Edge Chat team is committed to providing support and improving the app based on user feedback. Note that the app requires significant device storage and processing power, and performance may vary depending on your device's capabilities and the size of the models you use.