Z-Image AI
Z-Image AI — The ultimate creative suite powered by S3-DiT. Experience lightning-fast generation with Z-Image-Turbo and precise instruction-based modification with Z-Image-Edit.
Create & Refine
Discover how Z-Image AI transforms your creative workflow
1. Generate with Turbo
Start by using Z-Image-Turbo to generate high-quality images in milliseconds. Its 8-step inference allows for rapid idea visualization.
Generated in seconds with stunning
Impressive speed without compromising quality


2. Edit with Instructions
Switch to Z-Image-Edit to perfect your image. Use natural language commands to add, remove, or change elements without masks.
Instruction: 'Make the car red'
Instruction: 'Remove the background people'


3. Final Polish
Combine the speed of Z-Image-Turbo and the control of Z-Image-Edit to produce commercial-ready assets in record time.
Refine details
Export high-resolution image


Technology
Powered by S3-DiT Architecture
Z-Image AI leverages the revolutionary Single-Stream Scalable Diffusion Transformer (S3-DiT) to deliver state-of-the-art performance across generation and editing.
- Turbo Speed
- Z-Image-Turbo redefines efficiency. By distilling the massive 6B parameter Z-Image-Base model, it achieves high-quality results in just 8 sampling steps.
- Precise Editing
- Z-Image-Edit offers context-aware manipulation. It understands the physical properties of your image, ensuring edits blend seamlessly with the original lighting.
- Bilingual Support
- Both Z-Image-Turbo and Edit are trained on extensive English and Chinese datasets, providing superior text rendering and prompt understanding.
- Unified Core
- The S3-DiT architecture unifies our models, ensuring that the style and quality remain consistent whether you are generating or editing.
General FAQ
- What is Z-Image AI?
Z-Image AI is a platform featuring advanced generative models: Z-Image-Turbo for fast creation and Z-Image-Edit for precise modification, all built on Tongyi's S3-DiT technology.
- How fast is Z-Image-Turbo?
Z-Image-Turbo is designed for real-time performance. It can generate high-fidelity 1024x1024 images in approximately 8 sampling steps, significantly faster than traditional diffusion models.
- Can Z-Image-Edit remove objects?
Yes, Z-Image-Edit excels at object removal and replacement. You simply type an instruction like 'remove the tree', and the AI fills the gap intelligently.
- Is Z-Image-Base available?
The technology of Z-Image-Base serves as the foundation for Z-Image-Turbo. We currently offer the Turbo experience to provide the best balance of quality and speed.
- Is it free to try?
Yes, we provide free trial access to both Z-Image-Turbo generation and Z-Image-Edit tools so you can experience the power of S3-DiT.