ChatLlama is a native iOS & Mac app that provides a modern chat interface for interacting with the Ollama and OpenAI models. The app supports both OpenAI's models and locally hosted Ollama models, featuring a clean and responsive chat interface built on top of MessageKit.

Text Gen	Image Gen

• iOS 18.2+
• Xcode 16.0+
• For Ollama: Mac running Ollama
• For OpenAI: Valid API key

1. Clone the repository:

git clone https://github.com/yourusername/ChatLlama.git
cd ChatLlama

2. Open the project in Xcode:

open ChatLlama.xcodeproj

3. Install dependencies using Swift Package Manager:

   • MessageKit (4.0.0+)
   • MacPaw/OpenAI

4. Configure your LLM provider:

   • For OpenAI: Add your API key to Secrets.xcconfig
   • For Ollama: Follow the Ollama setup instructions below

5. Build and run the project (⌘+R)

1. Create a Secrets.xcconfig file if it doesn't exist
2. Add your OpenAI API key:

OPENAI_API_KEY = your_api_key_here

1. Install Ollama on your Mac by following the instructions at Ollama's website

2. Run your desired model locally:

ollama run llama3.2:latest

The app uses a flexible configuration system through LLMConfig.swift:

• Choose between OpenAI and Ollama providers
• Configure model settings
• Set custom endpoints and API parameters

Default configurations:

• OpenAI: Uses GPT-4o Mini model
• Ollama: Connects to http://localhost:11434 with Llama 3.2

ChatLlama follows an MVVM (Model-View-ViewModel) architectural pattern with clean separation of concerns:

• Models: Message and Sender types for chat data
• Views: UI components and ChatViewController for user interaction
• ViewModels: ChatViewModel containing business logic and state management
• Configuration: LLMConfig for LLM provider settings
• Networking: API services for provider communication

This project is licensed under the MIT License - see the LICENSE file for details

• MessageKit for the chat UI components
• Ollama for the local LLM infrastructure
• MacPaw/OpenAI for the OpenAI API client