ai-agent

gpustack: Scale AI Model Serving on GPU Clusters with Ease

Name: gpustack: Scale AI Model Serving on GPU Clusters with Ease
Author: gpustack

Deploy high‑performance LLM inference (vLLM, SGLang) on on‑demand GPU instances and get SSH access for rapid iteration. Ideal for founders building AI‑powered products. 5.8k+ GitHub stars

From the repo · “A GPU cluster manager for high-performance AI model serving (vLLM, SGLang) and on-demand SSH-accessible GPU instances.”

intermediate⏱ 1-2 hours💵 Free + LLM API costs

5,879 stars+41★ this week695 forksPythonGuide quality 8/10 · reviewed Oct 6Updated 10/23/2026100% free · open source

What it is

Manage a group of computers with good graphics cards to work together on AI tasks.

Real use case example

"A founder wants to build a high-quality chatbot for their e-commerce website. They create a virtual cluster using GpuStack, connect it to their preferred AI model, and start processing customers' queries in real-time. The founder can easily scale their cluster up or down as needed to handle changing traffic."

Best insideClaude DesktopCursor

When to use it

•When deploying large-scale AI models that require significant GPU resources
•When needing to orchestrate multiple AI models and inference engines
•When requiring a scalable and configurable GPU cluster for AI workloads

Ready-to-paste prompt

gpustack deploy --model resnet50 --engine vLLM --gpu-type NVIDIA_A100 --num-gpus 4

Heads up: Ensure you have a compatible NVIDIA GPU driver installed and configured on your system, as gpustack requires this to function correctly

Saves to your device

Use with Claude

New

Skip the builder — one click puts this in Claude, Cursor, Antigravity and more.

✅ Light setup

Installs with a command or two; your AI agent can do it for you.

Try it instantly — no install

Claude Code

mkdir -p ~/.claude/skills/gpustack && curl -fsSL https://workflowstacks.com/api/skills/gpustack/claude-skill -o ~/.claude/skills/gpustack/SKILL.md

Open in another AI app

Opens the app with this repo with the prompt ready to go — no copy-paste needed.

Connect the whole catalog (MCP)

claude mcp add --transport http workflowstacks https://workflowstacks.com/api/mcp

Adds a WorkflowStacks connector to Claude Code: search and load any skill here by chatting.

How gpustack: Scale AI Model Serving on GPU Clusters with Ease works

Codeflow

Free to inspect

Gpustack: Scale AI Model Serving on GPU Clusters with Ease is a very large Python project (~148k lines across 440 code files, plus 212 test files). Setup is light: installs like a normal app. Reading the code is optional. Last commit this month, Apache-2.0 license, has a test suite.

Size

Very large codebase

~148k lines · 440 code files · days to read — use, don't read

Setup

One-command install

Installs like a normal app. Reading the code is optional.

Runs on

Python

No API keys detected

Python 98%Shell 1%Jinja 1%

Where to start reading

1
README.md
Start here — what it does and how to install it
2
AGENTS.md
The instructions the AI actually follows
3
gpustack/main.py
Where the program starts running
4
pyproject.toml
Dependencies and the commands it exposes
5
gpustack/__init__.py
Inside gpustack/ — the main logic begins here

What's in each folder

gpustack/Core code — the actual logic412 files

docs/Documentation400 files

pack/Folder112 files

hack/Helper scripts36 files

charts/Folder21 files

docker-compose/Folder10 files

tests/Tests — proof it works300 files

static/Images & static assets47 files

READMEHas testsDocumentedCI checksApache-2.0 licenseUpdated this month

Use with Claude

New

Skip the builder — one click puts this in Claude, Cursor, Antigravity and more.

✅ Light setup

Installs with a command or two; your AI agent can do it for you.

Try it instantly — no install

Claude Code

mkdir -p ~/.claude/skills/gpustack && curl -fsSL https://workflowstacks.com/api/skills/gpustack/claude-skill -o ~/.claude/skills/gpustack/SKILL.md

Open in another AI app

Opens the app with this repo with the prompt ready to go — no copy-paste needed.

Connect the whole catalog (MCP)

claude mcp add --transport http workflowstacks https://workflowstacks.com/api/mcp

Adds a WorkflowStacks connector to Claude Code: search and load any skill here by chatting.

Quick Actions

Details

Creator

gpustack

Language

Python

Do this with a working workflow

Pages we wrote and tested that put a skill like this to work — importable n8n templates, outcome guides and Claude Desktop configs.

Free n8n template

AI Support Agent (Chat)

A chat agent that answers customer questions from YOUR business info — with conversation memory. It never invents policies; anything it doesn't know, it hands to a human.