FICHA · AUR

shimmy-gpu

Drop-in OpenAI API Replacement for Local LLMs - CUDA + Vulkan + OpenCL

  • local LLM API server
  • CLI
  • Server
  • Launchable
  • Runs in terminal
official+codex · reviewed · Jun 4, 2026 description in en

Description

Local LLM applications can expose an OpenAI-compatible API with GPU acceleration through CUDA, Vulkan, or OpenCL. This server is useful for developers testing AI tools against local models while using available graphics hardware. Prompts, model files, GPU memory pressure, and network binding should be managed deliberately.

How to run

shimmy

Commands: shimmy

Permissions

Permissions not analysed for this source yet.