Deploy on your GPU

A “Deploy on your GPU” button for your README

Projects that need a local, OpenAI-compatible LLM can link to a recipe: a list of models from the gpuOS catalog, as Ollama tags. Visitors sign up, connect their GPU and get the models deployed. Up to 10 models per recipe.

Add the button to a README

The link takes the models as Ollama tags, separated by commas: https://gpuos.si/deploy?models=<tag>,<tag>. Tags outside the gpuOS catalog are left out. This example deploys Qwen3 8B and BGE-M3.

Markdown

[![Deploy on your GPU with gpuOS](https://gpuos.si/badge/deploy.svg)](https://gpuos.si/deploy?models=qwen3:8b,bge-m3&ref=deploy-badge)

HTML

<a href="https://gpuos.si/deploy?models=qwen3:8b,bge-m3&amp;ref=deploy-badge"><img src="https://gpuos.si/badge/deploy.svg" alt="Deploy on your GPU with gpuOS" height="20"></a>