SmolChat
A local LLM chat app for Android that runs GGUF models on-device through llama.cpp. Open source under the Apache 2.0 license.
SmolChat is an Android app for people who want to chat with language models running on their own phone. It uses llama.cpp to run GGUF models on-device, including small language models. Inference stays on your device.
You choose the model files rather than relying on a single built-in model. The app lets you add and remove models, change their system prompts, and adjust response settings. That makes it useful for trying different models or giving one model different instructions for separate tasks.
Chat conversations retain their messages so the app can use them as context when you return. A separate task mode generates responses without saving those messages. This distinction suits people who want ongoing conversations alongside individual prompts they don't need to keep in chat history.
Responses support Markdown formatting and syntax highlighting for code, so the interface can display structured answers and code examples. SmolChat provides a mobile chat interface around GGUF inference rather than requiring users to work directly with llama.cpp.
The project is open source under the Apache 2.0 license. Its Android interface is written primarily in Kotlin, with llama.cpp handling model execution. The app is available through Google Play and as an APK from the project's releases.
