MLX Chat Example

A lightweight chat application demonstrating MLX integration for iOS and macOS. Built with SwiftUI, this example project shows how to implement both Large Language Models (LLMs) and Vision Language Models (VLMs) using MLX.

Features

🤖 LLM and VLM support with real-time text generation
📱 Cross-platform (iOS and macOS)
🖼️ Image and video input for vision models
💾 Efficient model caching and memory management
⚡️ Async/await based generation with cancellation support
🎨 Modern SwiftUI interface
📝 Comprehensive documentation and comments

Requirements

iOS 17.0+ / macOS 14.0+
Xcode 15.0+
Swift 5.9+

Dependencies

MLX: Core machine learning operations
MLXLMCommon: Common language model utilities
MLXLLM: Large language model support
MLXVLM: Vision-language model support

Project Structure

MLXChatExample/
├── Views/                # SwiftUI views
├── Models/               # Data models
├── ViewModels/           # Business logic
├── Services/             # MLX integration
└── Support/              # Utilities

Technical Overview

The project follows MVVM architecture with clear separation between UI and business logic. The core functionality is split into two main components:

MLXService

Core service handling all model operations:

Model loading and caching with memory management
Async text generation with streaming support
GPU memory optimization
Model state management
Handles both LLM and VLM model types

ChatViewModel

Business logic coordinator:

Manages chat state and message history
Handles generation lifecycle and cancellation
Coordinates media attachments for vision models
Performance metrics and error handling
Provides clean interface between UI and ML service

Architecture Highlights

Complete separation of UI and business logic
SwiftUI views with async/await integration
Modular design for easy extension

Documentation

The codebase is thoroughly documented with:

Detailed class and method documentation
Clear inline comments explaining complex logic
DocC documentation format

Markdown Support

This sample app renders markdown content using SwiftUI's native Text view by passing the content as a LocalizedStringKey:

Text(LocalizedStringKey(message.content))

Limitations and Alternatives

The default SwiftUI markdown rendering only supports standard markdown syntax. It does not support advanced features like tables and task lists that are available in GitHub Flavored Markdown (GFM).

For more comprehensive markdown support:

GitHub Flavored Markdown: Consider using the swift-markdown-ui library. However, be aware that this library currently has an unresolved issue with text selection, which is why it wasn't used in this example.
Enhanced Text Selection: If you're satisfied with standard markdown but want better text selection capabilities on iOS (instead of only being able to select and copy entire content block), consider combining:
- SelectableText for improved selection functionality
- MarkdownToAttributedString for markdown formatting

More discussion on this can be found on issue #297

Getting Started

Clone the repository
Install dependencies
Open in Xcode
Build and run

Contributing

This is an example project demonstrating MLX capabilities. Feel free to use it as a reference for your own projects.

Acknowledgments

MLX team for the core framework

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

Uh oh!