A native desktop client for interacting with Language Models (LLMs) built with Rust and Iced.

• 💬 Chat conversations with LLMs
• 📂 Conversation organization with folders
• 🔍 RAG (Retrieval Augmented Generation) support
• 🎛️ Multiple AI provider support by OpenAI-Api-Like (OpenAI, Ollama, Gemini, etc.)
• 📦 Preset management for different conversation settings
• 📊 Vector database integration
• 🖥️ Cross-platform (macOS, Windows, Linux)
• 🌙 Dark theme UI

Download pre-built binaries for your platform from the Releases page.

• Rust toolchain (stable, 2024 edition)
• Cargo

Clone the repository:

git clone https://github.com/pashaish/lm_client.git
cd lm_client

Build the project:

cargo build --release

Run the application:

cargo run --release

There are several build scripts available in the scripts directory:

• build-mac-arm.sh: Build for macOS ARM
• build-windows.sh: Build for Windows
• build-linux.sh: Build for Linux
• build-bundle.sh: Create application bundle

lm_client/
├── framework/           # Union functionality for src
├── modules/
│   ├── api/             # API clients for LLM providers
│   ├── database/        # Database implementation
│   ├── services/        # Business logic services
│   ├── types/           # Shared data types and DTOs
│   └── utils/           # Utility functions
├── src/                 # Main application code
│   ├── app/             # Main application state and view
│   ├── theme/           # UI theme definitions
│   └── widgets/         # Custom UI widgets
└── scripts/             # Build and utility scripts

The application follows an elm-like architecture using the Iced framework:

• State: Manages the application data
• Update: Handles events and updates state
• View: Renders UI components
• Subscription: Manages asynchronous events

The project uses a modular approach with workspaces to separate concerns.

Contributions are welcome!

See the todo.md file for planned features and improvements.

Short-term goals:

• Abort Chat Completion handling
• Notification system for errors and tips
• Custom error handling

Long-term goals:

• Custom markdown parser with text selection support
• PDF to text conversion for RAG
• Favorite models functionality

This project is licensed under the LICENSE file in the repository root.

• Iced - GUI library for Rust
• SQLite - Embedded database