Directory › Local AI › Llama pour Windows
> GEM LOCAL AI · PERSONAL AGENTS
Llama pour Windows
by ggml-org (llama.cpp)
A local AI chat on Windows 11, no command line: the app lives in your tray, downloads models in one click and opens with Alt+Space.
View on GitHub →What it does
- Pick a model from the list (or any GGUF from Hugging Face): one-click download, with memory estimated before you launch
- Alt+Space opens the chat over any app; the window hides instead of closing, so your conversation stays put
- No telemetry, models kept in the standard Hugging Face cache shared with llama.cpp; a ready-to-copy curl request to plug in your other tools
! Official app from the llama.cpp team, MIT-licensed code. No telemetry: your prompts stay on your PC. Two catches: the Alt+Space shortcut is hard-coded (it clashes with PowerToys Run, whose default it is), and if Smart App Control is on, Windows may block the llama.exe engine (unsigned). Information from the official repository (version 0.12.0) and tech press.
Install it
Requirements: Windows 11 (x64 or ARM64). If Smart App Control is on and blocks llama.exe, the server will not start.
https://github.com/ggml-org/Llama-Windows/releases
Llama-Windows.msixbundleAlt + Espace
The prompt for your agent
Paste it into your coding agent: it checks the requirements and installs it for you.
Install Llama pour Windows (https://github.com/ggml-org/Llama-Windows) on my machine.
1. Check the requirements: Windows 11 (x64 or ARM64). If Smart App Control is on and blocks llama.exe, the server will not start. If anything is missing, tell me before installing anything.
2. Follow the official method:
- see https://github.com/ggml-org/Llama-Windows/releases
- Double-click it to install (the app downloads llama.exe if you don't have one): Llama-Windows.msixbundle
- Click the tray icon, pick a model, then open the chat: Alt + Espace
3. Check that it works, then explain how to launch it and where to start.
Always review what your agent is about to run, especially install commands.