# Llama GUI

A lightweight local launcher and browser control panel for llama.cpp on Windows, macOS and Linux, covering installation, launching, chat, benchmarking and preset management.

Llama GUI is a local launcher and control panel for llama.cpp. It installs prebuilt llama.cpp releases for a chosen backend, such as CPU, CUDA, Vulkan, SYCL or ROCm, and launches the llama-server or llama-cli binary either from a beginner-oriented Quick Launch or a full Configure screen, all through a browser-based interface.

It provides a chat view with streaming Markdown, a focus mode and collapsed reasoning output, plus optional web search with no API key; benchmarking through built-in benchmark and perplexity tools with live stats; OpenAI-compatible API usage snippets; and management of launch presets, including two presets kept on standby for quick switching, Windows shortcuts, and in-app updates from GitHub releases. Five built-in themes meet accessibility contrast guidelines.

It installs with a one-line script on macOS, Linux or Windows, needs Python for the installer itself, and its supported prebuilt backends vary by operating system and architecture.

Website: https://github.com/thomas9120/LLama-GUI