Skip to content

WorkflowFreeIntermediate

Z-Image-Turbo Text to Image

Makes photo-quality images from a text prompt on your own computer with a fast open model

View original on GitHub

Install request

Your browser blocked copying. Select the text and copy it yourself.

Paste this into your AI agent (Claude Code, Codex or similar). It asks the agent to read the files and wait for your OK before installing.

Example output of the template: a photographic portrait of a woman with long curly hair at a harbour with pastel houses
Image: Comfy Org (MIT). Original file

What it does

This is an official ComfyUI workflow template for Z-Image-Turbo, an open image model from Alibaba's Tongyi Lab. You write a prompt in English or Chinese and it generates an image locally, with no per-image fees. ComfyUI offers to download the three model files (about 19 GB) the first time you open the template. It suits people with a gaming-class graphics card who want to make images without an online subscription.

The model is small enough for a 16 GB consumer graphics card and can draw readable English and Chinese text inside images.

Before you start

  • comfyui, on your own server or the platform's cloud.
  • ComfyUI installed: Comfy Desktop (Windows, macOS, Linux), Portable or a manual install
  • A graphics card with about 16 GB of memory, per the model makers' figure
  • About 19 GB of free disk space for the model files
  • Free. Nothing to pay.

How to start

  1. Open comfyuiYour own server or the platform's cloud.
  2. Import the templateDownload it from the creator's page and import it or follow the official guide. An AI agent can walk you through it with the install request.
  3. Run it once with test dataConnect your accounts, try it on something small and check the result.

Steps follow the creator's instructions.

Stuck setting it up?

Tell us what happened. We confirm scope and price before any paid work.

Get help
View original

Install request

Your browser blocked copying. Select the text and copy it yourself.

Paste this into your AI agent (Claude Code, Codex or similar). It asks the agent to read the files and wait for your OK before installing.