AI Models
What each model is good at, where it differs from the others, and what it costs to run.
The first Qwen image model from Alibaba. Strong type in the picture, plus one attached picture to edit or to transform.
Credits come with a Generator plan. The exact cost is shown on the Generate button before anything is spent.
Qwen is Alibaba's first Qwen image model, Qwen-Image. Use it when the picture has to render real text, in English or Chinese, or when one picture you already have needs a precise change.
The Qwen team describes Qwen-Image as a foundation model for complex text, including multi-line layouts, and for edits that keep both the meaning and the look of the picture. Qwen-Image-Edit extends that to changing, removing, or restyling parts of a picture, and to editing the words already in it.
Write the picture in the prompt. To change one picture, open Qwen Edit. To transform one picture, open Qwen Image. Each takes a single picture. There is no Enhance switch.
Qwen Image 2 is the later model, and it edits up to three pictures in one pass. This model takes one. Z-Image is another Alibaba image model, and on Visxy it is text only.
Qwen Edit changes one attached picture. Qwen Image transforms one attached picture. Both sit beside this model. A new picture from text alone starts here.
No. One picture. Up to three is Qwen Image 2.
The Qwen team describes high-fidelity text in both. Ask for the words in the prompt. This model does not edit an existing picture's type unless you use Qwen Edit.
No.
