FICHA · AUR

llama.cpp-vulkan

Port of Facebook's LLaMA model in C/C++ (with Vulkan GPU optimizations)

  • cli-tool
  • TERMINAL
  • AI
  • GPU
  • Launchable
  • Runs in terminal
official+codex · reviewed · May 29, 2026 description in en

Description

Large language models can run locally with a C and C++ inference engine accelerated through Vulkan-capable GPUs. It helps developers, researchers, and advanced users test compatible LLaMA-style models without sending prompts to an online service.

Model files are separate and can be large. Local AI output can be wrong, biased, or unsafe for decisions that require expertise, and prompts may still expose private data to anyone with access to the machine.

How to run

llama-cli

Commands: llama-cli

Permissions

Permissions not analysed for this source yet.

Project health

Active

Last commit: