[Workflow] Z-Image Turbo: Fast Image Generation in Seconds — v1.0

RUN THIS WORKFLOW NOW ON FLOYO

HOW IT WORKS

Step 1. Write your prompt Describe the image you want. Name the subject, the lighting, and the setting. Z-Image Turbo is strong at photorealism and at rendering short text inside the image, so spell out any words you want to appear. Works great with: photorealistic scenes · product shots · images with text

Step 2. Hit run The model and settings are already dialled in. For a first run, change nothing, hit run and wait a few seconds.

Step 3. Download your image You get back a 1024 x 1024 PNG. It drops straight into any design or editing tool. Ready for: Photoshop · Figma · Canva · any editor

First time? Leave every setting as-is. The defaults (1024 x 1024 · 9 steps · CFG 1 · random seed) are the right starting point for almost everyone.

RECOMMENDED SETTINGS

Quick-start guide. Find the goal that matches yours and copy the settings.

Standard generation (most people) Start here : 1024 x 1024 · 9 steps · CFG 1 · euler · random seed. The right starting point for almost everyone.

Want text in the image to come out clean : Be explicit in the prompt about the exact words and where they sit. "A sign that reads 'OPEN' in bold red letters" reads better than "a sign."

Reproduce a result you liked : Lock the seed to the number that produced it. Same seed and same prompt gives you the same image again.

Want a batch of variations : Keep the seed on random and raise the batch size. Each run gives you a fresh take on the same prompt.

Portrait or landscape framing : Change the width and height on the latent. 832 x 1216 for portrait, 1216 x 832 for landscape. Stay near a 1-megapixel total for the cleanest result.

Result looks over-cooked or unnatural : Keep CFG at 1. Z-Image Turbo is a distilled turbo model and is tuned to run without heavy guidance. Pushing CFG higher tends to harden edges and wash out detail.