Your first local Qwen setup
Install Ollama, download Qwen3, and start a conversation on your own machine.
Before you begin
- Difficulty
- Beginner
- Time required
- Not yet tested
- Software
- Ollama
- Hardware
- Not yet tested
- Required VRAM
- Not yet tested
- Dependencies
- See official setup guide
The workflow
Install Ollama
Download the installer for your operating system from ollama.com. Follow its installation instructions.
Download and run Qwen3
Open a terminal. The first run downloads the model; storage and memory needs depend on the selected package.
ollama run qwen3:8bCheck memory and responsiveness
Start with a short prompt. Monitor memory and GPU utilization before increasing context or running other applications.
When it doesn’t go to plan
Close other GPU applications, lower the context, or select a smaller model. Hardware requirements have not yet been tested here.
The model behind this workflow
Qwen3 8BKeep exploring
Build a local coding stack
Connect Qwen2.5 Coder to a local Ollama endpoint and your coding tools.
Make images with FLUX & ComfyUI
Follow the official node workflow for FLUX.1 schnell, with model placement and a first render.
From a prompt to a local video
Use the official Wan2.1 ComfyUI workflow to explore text-to-video generation.