Chatty-Mini

Fractal Media Infrastructure

About this app

Chatty-mini: Your Personal, Private AI Workspace

Experience the power of local generative AI right in your pocket. Chatty-mini is a lightweight, portrait-first assistant designed to run GGUF models directly on your device. Whether you need an offline companion for creative writing, a fast assistant for your notes, or a secure space for your private documents, Chatty-mini puts you in total control.

Why Choose Chatty-mini?
Most AI apps force your data into the cloud. Chatty-mini is built for privacy—your prompts and files stay on your device by default. It is specifically optimized for small-screen Android devices, offering a clean, distraction-free interface for one-handed operation.

Core Features

Local Inference: Run GGUF models locally. No tracking, no mandatory cloud sync.

Small-Phone Optimized: A refined, compact UI designed to minimize clutter and maximize chat space on devices like the Samsung A21s and similar handsets.

Model Versatility: Import your own GGUF files and assign specific models to "Main AI" or "Bookkeeper" roles to balance speed and intelligence.

Integrated Sandbox: Manage your project files directly within the app. Open, edit, and save .md, .txt, and .json files.

Advanced Persona Switching: Save and toggle between character profiles to instantly adapt your AI's tone and expertise to your current task.

Optional Cloud Power: Need more reach? Use our secure, optional integration for OpenAI, Anthropic, Gemini, xAI, or DeepSeek. Your API keys are encrypted locally using the Android Keystore, and cloud usage is strictly user-directed—there is no hidden, automatic fallback.

Developer-Friendly: Chatty-mini is free and open-source software (FOSS). Built with Flutter and a bundled native runtime to provide a seamless out-of-the-box experience.

Privacy-First Approach
Your data is your own. We do not transmit your local chats or files to any server. When you use optional cloud features, you decide exactly when and where your data is sent.

Open Source
Chatty-mini is released under the GNU Affero General Public License v3.0 (AGPLv3). We believe in transparency and user freedom. You can study, modify, and redistribute the software, and we invite the community to participate in its development.

AI Disclosure
Chatty-mini provides access to generative AI capabilities. All content generated by the AI is produced based on the models loaded or configured by the user. Users are responsible for verifying the accuracy and appropriateness of all AI-generated content.

Getting Started with Local AI
Chatty-mini is an engine for your own GGUF models. To begin, you will need to download a GGUF-format model file.

*Beginner Tip: If you are new to local LLMs, we recommend searching for "GGUF" models on Hugging Face. For small phones with limited RAM, look for sub-1B parameter models or "tiny" variants to ensure smooth, responsive performance.

Chatty-mini requires no setup of external runtimes; it comes fully bundled and ready to go.

What’s new

Initial Release