Skip to main content

​
About the Provider

Alibaba, through its Tongyi-MAI research division, is the technology provider behind Z Image Turbo (Qwen Tongyi MAI). Alibaba is a major Chinese tech and cloud computing company that develops and open-sources advanced AI models and tools to support innovation, multimodal AI research, and broad integration in applications.

​
Model Quickstart

This section helps you quickly get started with the Tongyi-MAI/Z-Image-Turbo model on the Qubrid AI inferencing platform. To use this model, you need:
  • A valid Qubrid API key
  • Access to the Qubrid inference API
  • Basic knowledge of making API requests in your preferred language
Once authenticated with your API key, you can send inference requests to the Tongyi-MAI/Z-Image-Turbo model and receive responses based on your input prompts. Below are example placeholders showing how the model can be accessed using different programming environments.
You can choose the one that best fits your workflow. 
import requests

url = "https://platform.qubrid.com/api/v1/qubridai/image/generation"

headers = {
  "Authorization": "Bearer <QUBRID_API_KEY>",
  "Content-Type": "application/json"
}

data = {
  "model": "Tongyi-MAI/Z-Image-Turbo",
  "positive_prompt": (
      "A retro 1980s synthwave album cover. A grid landscape leading to a setting purple sun "
      "in the distance. A chrome sports car driving away. The text 'Qubrid is Qool' is written "
      "in a metallic chrome script font with neon pink outlines floating in the sky. "
      "CRT monitor effect, grain, vibrant neon colors."
  ),
  "width": 1024,
  "height": 1024,
  "steps": 9,
  "cfg": 0,
  "seed": 42
}

response = requests.post(url, headers=headers, json=data)

if response.status_code == 200:
  with open("generated_image.png", "wb") as f:
      f.write(response.content)
  print("Image saved to generated_image.png")
else:
  print(f"Error {response.status_code}: {response.text}")

​
API Generated Response

​
Example Generated Image

Prompt: A professional product photography shot of a matte black coffee bag sitting on a wooden table. The bag clearly features the text ‘QUBRID ROAST’ in gold foil stamping. Beside the bag is a white ceramic cup with latte art. Morning sunlight, sharp focus, 8k resolution, commercial advertisement style.

​
Model Overview

Z Image Turbo is a distilled variant of the Z-Image family designed for fast and efficient image generation.
  • It is built on a 6B-parameter foundation model and optimized to achieve strong image quality with low inference cost.
  • Z Image Turbo supports photorealistic image generation, bilingual text rendering (English and Chinese), and robust instruction adherence.
  • The model is optimized for low latency inference with a fixed single-image output and supports a wide range of image resolutions.
  • It also provides optional prompt enhancement with reasoning support through the prompt_extend parameter.

​
Model at a Glance

FeatureDetails
Model IDTongyi-MAI/Z-Image-Turbo
ArchitectureLatent Diffusion Turbo model based on Z-Image framework designed for high-speed image generation
Model TypeImage Generation
Model Size6B params

​
When to use?

You should use Z Image Turbo if you need:
  • Fast image generation with low inference latency.
  • Photorealistic image outputs with strong aesthetic quality.
  • Support for bilingual text rendering in English and Chinese.
  • Flexible image resolution control within a defined pixel range.
  • Optional prompt rewriting and reasoning through prompt enhancement.
Z Image Turbo is suitable for production inference scenarios where efficiency, consistency, and predictable output behavior are required.

​
Inference Parameters

Parameter NameTypeDefaultDescription
Widthnumber1024Generated image width.
Heightnumber1024Generated image height.
Inference Stepsnumber9Number of denoising steps (Turbo models require fewer steps).
Guidance Scalenumber0Classifier-free guidance scale.
Seednumber42Random seed for reproducibility.
Negative Promptstring—Optional negative prompt.

​
Key Features

  • High-speed image generation using a distilled 6B-parameter model optimized for low latency
  • Photorealistic and stylized outputs with strong prompt adherence
  • Bilingual text rendering with reliable English and Chinese typography
  • Resolution flexibility supporting 512×512 up to 2048×2048 total pixels
  • Deterministic outputs with seed control and fixed single-image generation
  • Production-ready inference with predictable cost and performance

​
Summary

Z Image Turbo is a fast, efficient image generation model built for production use. It is distilled from a larger foundation model to reduce inference cost while maintaining visual quality. The model excels at photorealistic images, stylized artwork, and text rendering. It supports flexible resolutions and deterministic outputs via seed control. Z Image Turbo is ideal for real-time and large-scale image generation workloads. For a comprehensive overview of Z-Image-Turbo
visit Qubrid’s Official Medium Guide.