AI Image Generator

The official 6B text-to-image Turbo variant: fixed 8-step generation for faster inference and high visual quality. Supports Chinese/English/mixed prompts, with a focus on prompt following and native text rendering.

Generate images with WAN 2.5. Choose a resolution, tune optional params, and create high-quality results.

A balanced, general-purpose model for fast creation. Great for concepts, illustrations, and stylized outputs.

A cost-effective Pro option for everyday use. Great for fast iterations, solid details, and consistent quality presets.

Maximum fidelity and layout control for premium product shots, editorial portraits, posters, UI mockups, and infographic outputs with dense text.

Generate high-fidelity visuals with retrieval-aware reasoning, instruction control, and multi-step composition.

A strong creative model for aesthetics and stylization. Works well for posters, scenes, and rich lighting.

Fast and cost-efficient. A great default for high throughput generation and quick iteration.

Give it an image and a clear instruction, and it edits what you specify while preserving identity and details: outfit/hairstyle changes, object add/remove, material swaps, text edits, lighting/viewpoint changes, and more.

Remove backgrounds for product images and portraits. Export with a transparent background or solid color.

Increase resolution while preserving details. Great for making generated images sharper and print-ready.

Decompose an image into editable RGBA layers, then recolor, replace, or revise a single layer while keeping structure.