Alibaba: Z-Image-Turbo

Z-Image Turbo is a super fast text-to-image model of 6B parameters developed by Tongyi-MAI.

z-image-turbo

AvailableAlibabaImageSync

Price: From $0.015
Modalities: Image

About Z-Image-Turbo

Z-Image-Turbo is the distilled, speed-oriented version of Tongyi-MAI's 6-billion-parameter Z-Image model, released by Alibaba under Apache 2.0. It produces an image in 8 function evaluations, with sub-second latency on enterprise GPUs, and renders English and Chinese text well. It trades the guidance control and diversity of base Z-Image for quick, production-ready output.

Where it works well

Eight inference steps give very fast turnaround, which suits interactive tools and high-volume queues.
It renders both English and Chinese text accurately inside images.
Photorealistic output and strong instruction following hold up despite the short sampling schedule.

When to choose another model

It does not offer the classifier-free guidance and negative-prompt control of base Z-Image.
Outputs vary less between seeds than the base model, so exploration is narrower.
It generates from text only and cannot edit an uploaded picture.

Getting started

Create API key
Create a key in Console, then use it with every model on the platform.

Send your first request

Copy the example for your language and run it against the endpoint.

Text to imageTokenLab endpoint

POST/v1/images/generations

curl -X POST "https://api.tokenlab.sh/v1/images/generations" \
  -H "Authorization: Bearer sk-xxx" \
  -H "Content-Type: application/json" \
  -d '{
  "operation": "text-to-image",
  "model": "z-image-turbo",
  "prompt": "A minimalist product photo of matte black headphones on a soft blue background."
}'

Pricing

The Verified price applies to Verified, which costs less on most models. The Official price is the model maker's published price and applies to the more reliable Official route. Auto bills the route that completes the request.

prompt_extend · false

per image

Official price: $0.015
Verified price: —
Discount: —

prompt_extend · true

per image

Official price: $0.03
Verified price: —
Discount: —

	Official priceper image	Verified priceper image	Discount
prompt_extend · false	$0.015	—	—
prompt_extend · true	$0.03	—	—

Usage & activity

Success rate is the share of requests that completed. Latency is how long a full response takes; P95 means 95% of requests finished within that time.

Usage & availability

Last 24 hours

Requests
Success rate
P95 latency
Total tokens

Model performance

Last updated Oct 11, 2026, 04:33 AM UTC

Metrics appear once privacy and data volume thresholds are met.

Data is based on aggregate user requests, excluding status checks.

Open in Console

Open Z-Image-Turbo in Console with a prompt ready to edit or send.

Help me create an image with z-image-turbo using text-to-image at /v1/images/generations. Show the result, cost, and any limits I should know.

Use cases

Best for

Text to Image

Interactive image tools
Give users a result within a second or two in a chat, a design widget or a live preview.
Bulk asset generation
Fill a catalogue with illustrative images quickly when each one needs only light review.
Bilingual thumbnails
Make cover images with a short English or Chinese title baked into the picture.

Prompt examples

Photo of a street vendor selling mangoes at noon, shallow depth of field, a sign reading '新鲜芒果'.

Minimalist app icon of a paper plane in coral and navy, flat vector look.

Close-up portrait of an elderly fisherman, natural light, weathered hands.

FAQ

How fast is Z-Image-Turbo?

The model card states 8 function evaluations and sub-second latency on enterprise-grade GPUs. Actual response time through the API also depends on queue time and image size.

What is the difference between Z-Image-Turbo and Z-Image?

Turbo is distilled with Decoupled-DMD and DMDR for 8-step generation. Base Z-Image takes 28 to 50 steps and keeps guidance, negative prompts and greater diversity, which also makes it the one to fine-tune.

Can it write text in images?

Yes, in English and Chinese. Tongyi-MAI highlights accurate rendering of complex text in both languages as a distinguishing feature, which is unusual for a model that finishes in eight steps.

Is it open weights?

Yes. It is on Hugging Face under Apache 2.0, at 6 billion parameters, and the card says it runs within 16 GB of VRAM on consumer hardware.

How much does Z-Image-Turbo cost?

On TokenLab, Z-Image-Turbo costs $0.015-$0.03 per image. The pricing table above shows the full breakdown. Rates depend on the billing unit, specification, and usage. Compare matching conditions in the model's detailed pricing; a single rate does not determine the total cost.

Which endpoint should Z-Image-Turbo use?

Use https://api.tokenlab.sh/v1/images/generations for Z-Image-Turbo. The request example below shows the matching code shape.

Which operations does Z-Image-Turbo support?

Z-Image-Turbo supports Text to image. Select an operation above to see its endpoint and request example.