Qwen-Image: text in the image, and editing
Qwen-Image is a family of visual models from the Qwen ecosystem, noted for rendering legible text in the image, with an editing branch, Qwen-Image-Edit. What it brings to posters, packaging and signage, and its place in Nova since June 2026.
Qwen-Image is a family of image generation models developed in the Qwen ecosystem, designed with particular attention to text understanding and typographic rendering: legible words in the image, where many models fail. A dedicated branch, Qwen-Image-Edit, transforms an existing image by instruction. At Axiris, Qwen-Image 2512 has been one of Nova's engines since 28 June 2026; what the Qwen team announces and what we observe are documented separately.
Image generation and editing in the Qwen ecosystem
Qwen-Image is a family of visual models developed in the Qwen ecosystem. It is built around image generation and pays particular attention to text understanding and typographic rendering. A branch dedicated to editing transforms existing images.
Text in the image
Producing legible text in an image remains a problem for many models: distorted letters, invented words, unstable typography. Qwen-Image was presented with specific work on rendering complex text in visuals.
That capability matters for posters, graphic compositions, packaging, signage, advertising and layouts that carry text. It does not exempt you from the Axiris rule for anything that has to be exact: a logo or a final title is placed to the pixel, as a layer, on top of the render, never reinterpreted by a model. Generated text serves exploration and mock-ups; the final text is controlled.
Qwen-Image-Edit
The editing branch starts from an existing visual and applies instruction-guided transformations: changing an element, a light, a setting, while keeping the rest. It is one of the families that can edit, alongside FLUX.2, Gemini Image and GPT Image. See Image editing models.
This approach fits directly into the way Axiris builds an image, step by step, keeping what is validated.
Qwen-Image in Nova
Qwen-Image 2512 has been one of the engines available in Nova since 28 June 2026, along with Krea 2: Nova grows: Krea 2 and Qwen-Image 2512. Several engines under one direction means several ways of serving the same visual intention, not more chance.
Qwen in a workflow
Depending on the versions and tools available, Qwen plays a part in the chain generation → editing → composition → correction → render, locally or in ComfyUI. What the Qwen team announces and what we observe when using it are documented separately, under known conditions: Comparing AI image models. For a given job, which one to choose: How to choose an AI image model?.
Official sources
Contact request: a first conversation, free and with no commitment. Every practice is different; it is by listening to you that we propose what suits you.