Skip to main content
TikTomato
Sign in

Z-Image Turbo image generator

Run Tongyi's Z-Image Turbo from your browser. Write a prompt and get an image quickly, for 1 credit. Free credits when you sign up, nothing to install.

Reference images (optional) · Up to 4 images, JPG / PNG / WebP, 5 MB each

Z-Image Turbo is text-to-image only and has one output size (about 1 megapixel). To edit a photo or generate at 2K or 4K, use the standard image generator.

About Z-Image Turbo

Z-Image Turbo is an open-source image generation model from Alibaba's Tongyi-MAI team, released in November 2025 under the Apache 2.0 license. It has 6 billion parameters and is built on what the team calls a Single-Stream Diffusion Transformer.

Turbo is the distilled version of Z-Image: according to its model card it needs only 8 sampling steps per image, which is why it is fast and cheap to run. That is also why an image costs only 1 credit here.

What Z-Image Turbo is good at

The strengths listed on the Z-Image Turbo model card.

Photorealistic images

The model card puts photorealistic generation first: portraits, street scenes, product-style photos.

English and Chinese text

It renders text in both English and Chinese, so signs, posters and labels with Chinese characters come out readable.

Follows the prompt

The model card describes its instruction following as robust: what you ask for and where you place it tends to hold.

Fast

Eight sampling steps instead of the usual dozens. That is why images come back much faster here than on our standard generator.

Images generated with Z-Image Turbo on this page

Each of these was generated on this page. The caption is the exact prompt sent to the model.

Neon noodle shop sign with Chinese and English text on a rainy street, generated with Z-Image Turbo
“A neon sign above a small noodle shop on a rainy night street. The sign reads "番茄面馆" in large red Chinese characters, with "TOMATO NOODLES" in smaller white English letters below it. Wet pavement reflections, steam from the doorway, photorealistic, 35mm lens.”
Photorealistic portrait of a freckled woman with curly red hair and a yellow scarf, generated with Z-Image Turbo
“Close-up portrait of a young woman with freckles and curly red hair, wearing a mustard yellow knitted scarf, soft window light from the left, shallow depth of field, natural skin texture, shot on an 85mm lens, photorealistic.”
Farmers market vendor handing a paper bag to a customer, generated with Z-Image Turbo
“A busy morning farmers market stall piled with tomatoes, peppers and fresh herbs. An elderly vendor in a green apron hands a paper bag to a customer. Golden sunlight, candid documentary photograph, rich colors.”
White lighthouse at dawn with three seagulls and a red rowboat, generated with Z-Image Turbo
“A white lighthouse on a rocky headland at dawn, three seagulls in the sky, a small red rowboat in the lower left, photorealistic”

How to prompt Z-Image Turbo

What follows from how the model works.

Describe what you want, not what to avoid

The Turbo model does not support negative prompts. Instead of "no people", describe the empty street you want to see.

Put text in quotes

Put the exact words that must appear in quotes and say where they go. Chinese works as well as English.

Be specific

Subject, setting, light, lens, style. Prompts can be up to 1,000 characters, and the model uses the detail you give it.

Try several

At 1 credit per image and a short wait, the quickest way to a good result is to generate a few and adjust the prompt.

Z-Image Turbo FAQ

How much does an image cost?

1 credit per image, the lowest price of any tool on TikTomato. New accounts get free credits, so you can try it without paying.

How long does an image take?

Much less than on our standard generator. The exact time varies with load, and the first image after a quiet period can take longer.

What size are the images?

About 1 megapixel. A 16:9 image is 1280x720 pixels, 4:3 is 1152x864. You can choose 1:1, 4:3, 3:4, 16:9 or 9:16. For something larger, run the result through our image upscaler.

Can I edit an existing photo or use reference images?

Not with Z-Image Turbo: it is a text-to-image model. Our standard image generator takes up to 4 reference images.

How does it compare with GPT Image?

Z-Image Turbo is a much smaller model built for speed and cost. It is a good fit for quick ideas, drafts and photorealistic scenes. For complex compositions, long passages of text or editing photos, GPT Image on our standard generator is the stronger choice.

Can I use the images commercially?

The model is released under the Apache 2.0 license, which does not restrict commercial use of its output. You are responsible for what you create and how you use it.

What if my request is declined?

Prompts and results are checked for sexual and other sensitive content, and flagged requests are declined. Rephrase the prompt and try again. Credits for declined or failed requests are refunded.