FICHA · AUR

ollama-cuda-git

Create, run and share large language models (LLMs) with CUDA

  • llm-runtime
  • CLI
  • SERVICE
  • AI
  • Launchable
  • Background service
official+codex · reviewed · Jun 2, 2026 description in en

Description

Local language models can run through a development build of Ollama with CUDA acceleration. This helps users test NVIDIA GPU inference and newer upstream behavior.

It can consume large GPU memory and expose a local API. Confirm driver compatibility, model licenses, and prompt privacy before using it for real work.

How to run

ollama

Commands: ollama

Permissions

Permissions not analysed for this source yet.