LlamaStation v0.9

First public release 🎉

A GUI for running local AI models via llama.cpp — no command line needed.

What's included

• Full llama.cpp parameter control with per-model profile saving
• Real-time VRAM meter (multi-GPU support)
• Built-in model downloader from Hugging Face
• OpenAI-compatible API
• Dark/Light mode, English/Spanish UI
• Server watchdog with auto-relaunch
• Multiple backends: Official llama.cpp, TurboQuant, AtomicChat, BeeLlama

Installation

1. Download LlamaStation-v0.9.zip and extract the folder
2. Run LlamaStation.exe
3. Click Update backend to download llama.cpp automatically
4. Download a model and start chatting

Voice Mode (XTTS + Whisper) is not included in the .exe — run from source if you need it.