oxproxion

oxproxion is a versatile and user-centric Android chat application designed to interact with various Large Language Models (LLMs). It provides a seamless interface for managing conversations, customizing bot personas, and saving chat histories.

✨ Features

🤖 Multi-Model Support: Switch between different LLM bots and models.

🤖 Chat with OpenRouter models or locally served models using Ollama, LM Studio, llama.cpp, MLX LM or Hermes Agent.

💬 Chat Interface: A clean and intuitive interface for conversing with AI models.

💾 Save & Load Chats: Save your text chat sessions and load them later to continue the conversation.

📤📥 Import & Export: Easily import and export your chat histories.

⚡ Streaming or Non-Streaming Responses: You choose!

🖼️ Chat with Images: With models that support it!

🛠️ Tools:

Long-Press the Tools button in the options menu to enable/disable specific functionality:

• Read File
• Make File
• List Files in Workspace
• Make Timer
• Make Alarm
• Delete Files
• Make Calendar Event

🎨 Image Generation: With models that support it!

✍️ System Message Customization: Create, edit, and manage a library of system messages to guide the AI's behavior and persona.

🧠 Reasoning: Choose settings for reasoning models.

⚙️ Presets: Create, edit, and manage a library of presets to apply a combination of models, system messages, and options with one click. Presets are also exposed as system share targets.

🔤 Font: Customize the display font for optimal readability in the main chat screen.

🎙️ Conversation Mode: Speak to the AI and have responses read out loud automatically.

🔗 OpenRouter Integration:

  • Tap icon in model list to open the OpenRouter model's info in your browser.
  • Long-press the API key icon to view your remaining OpenRouter credits.

🔗 Ollama, LM Studio, llama.cpp, MLX LM and Hermes Agent Integration:

  • It now also allows chat with Ollama, LM Studio, llama.cpp, MLX LM or Hermes Agent models served on your LAN.

📋 Effortless Copying: Press any avatar to copy the corresponding response to the clipboard. Long-press the response avatar to copy in Markdown.

📄 On-Device PDF Export: Make a PDF of the whole chat with the PDF button in the options area. Or press the PDF icon/button on the response to generate a PDF of that specific response; All created locally using Android's native tools.

🛠️ Built with Modern Tech: 100 % Kotlin, leveraging Jetpack libraries, Coroutines for asynchronous tasks, and Ktor for networking.

Screenshots

Here are some screenshots of the app in action:

Main Screen

A view of the app's primary interface.

Image Generation

Image Generation

System Messages

Custom System Messages

SOTA Models Galore!

Add Models to your delight!

Markdown

Markdown delight!

Options with Extended Dock on and Reasoning on

Customize your experience in the options menu.

Your List

Your Model List

Vision Model

Vision Model

Advanced Reasoning Settings

Advanced Reasoning Settings

Presets

Presets Screen

Ollama, LM studio and llama.cpp Models

LAN Screen

️ Technologies Used

• UI: Android Jetpack, Material Components, View Binding
• Architecture: MVVM (ViewModel, LiveData/StateFlow)
• Asynchronicity: Kotlin Coroutines
• Database: Jetpack Room for local storage
• Networking: Ktor HTTP Client
• Markdown Rendering: Markwon

Getting Started

To build and run the project, follow these steps:

1. Clone the repository:

   git clone https://github.com/stardomains3/oxproxion.git
   

2. Open in Android Studio: Open the project in the latest version of Android Studio.
3. Build the project: Android Studio should automatically sync the Gradle project. Click the "Run" button to build and install the app on an emulator or a physical device.

⚙️ Configuration

To use with OpenRouter models, this application requires an API key from OpenRouter.ai to function. You will need to create an account, add credits, and obtain your API key.

Once you have your key, you can add it to the app's settings by pressing the "OpenRouter API Key" button.

Alternatively, you can run this locally with Ollama, LM Studio, llama.cpp, MLX LM or Hermes Agent.

To use these, you must host the model servers on your own LAN. Note: This application does not provide setup guidance for local infrastructure. You will need to configure the following yourself:

Server Setup: Read the official documentation for your chosen server to configure it correctly.

💡 Tip: Don't know where to start? Ask an AI to generate the specific configuration steps for your OS and hardware.

Network Configuration: Adjust your network environment, including firewall rules, router settings, and port forwarding.

Connection: Ensure your LAN allows traffic between the app and the model server.

Once you have your properly configured LAN endpoint for the server enter it at Settings->"LAN Settings" button.

Disclaimer: This project is not affiliated with, endorsed by, or sponsored by OpenRouter.ai in any way.

• This open-source app is provided 'as-is' without any warranty, express or implied. Use at your own discretion.
• Review OpenRouter's and its model providers' pricing, privacy, and logging policies on their websites.
• oxproxion app does not have trackers, analytics, nor ads.
• This app is intended for use by persons 18 and over.
• Chats are stored unencrypted on device only.

If you find this app useful, please consider supporting its developer: Buy Me a Coffee ☕

License

This project is licensed under the Apache License 2.0. See the LICENSE file for details.