Skip to main content

Examples

End-to-end walkthroughs for running AI workloads on Hyperstack. Each example starts from a GPU virtual machine or Kubernetes cluster and ends with a working setup you can build on.

Ollama

Run open-source LLMs on a GPU virtual machine with a REST API and OpenAI-compatible endpoint.

Open WebUI

Add a browser-based chat interface to an Ollama model running on a GPU virtual machine.

vLLM

Serve open-source LLMs behind an OpenAI-compatible API on a single GPU virtual machine.

Multi-GPU vLLM

Serve a large model across multiple GPUs on one virtual machine with vLLM tensor parallelism.

LLM on Kubernetes

Serve a large language model across a multi-node Kubernetes cluster with vLLM.

Fine-Tuning LLMs with Unsloth

Fine-tune Qwen3-8B with QLoRA on a GPU virtual machine using Unsloth.

ComfyUI

Generate images with FLUX.1-schnell and node-based workflows on a GPU virtual machine, over an SSH tunnel and an HTTP API.

Jupyter Notebooks

Deploy and run Jupyter Notebooks on GPU virtual machines.

Docker

Use Docker containers on Hyperstack virtual machines.

Ephemeral Drive Mounting

Customize ephemeral drive mounting behavior using cloud-init.