Introducing Qwen-Image: Novel Model in Image Generation and Editing

Alibaba released Qwen-Image, a novel image generation foundation model that achieves significant breakthroughs in complex text rendering and precise image editing.

The model can render intricate texts with high precision in generated images.
It’s an ideal foundation model to develop creative visual content, laying foundation for developing AI-driven applications

Alibaba released Qwen-Image, a novel image generation foundation model that achieves significant breakthroughs in complex text rendering and precise image editing. A dense model with 20 billion parameters, it achieves remarkable performance across a wide range of image generation and editing tasks, establishing itself as a leading model in the field.

The model is now open sourced on Hugging Face, GitHub and Alibaba’s open-source community ModelScope, and accessible on Qwen Chat under “Image Generation” model. The full technical report is also available online.

Watch the video to meet Qwen-Image.

Through innovative approaches such as comprehensive data engineering, progressive learning strategies, enhanced multi-task training paradigms, and scalable infrastructure optimization, Qwen-image delivers exceptional precision in rendering intricate text within generated images. It excels in challenging scenarios involving multi-line layouts, paragraph-level semantics, and fine-grained visual details. The model also demonstrates superior performance in consistent image editing, effectively preserving both semantic integrity and visual realism throughout the editing process.

[Prompt: Bookstore window display. A sign displays “New Arrivals This Week”. Below, a shelf tag with the text “Best-Selling Novels Here”. To the side, a colorful poster advertises “Author Meet And Greet on Saturday” with a central portrait of the author. There are four books on the bookshelf, namely “The light between worlds” “When stars are scattered” “The silent patient” “The night circus”]

Qwen-Image achieves remarkable advances in two key areas: generating high-quality, stylistically diverse images from complex textual prompts, and enabling context-aware image editing. Its editing capabilities include style transfer, text editing, background replacement, object addition, removal, or substitution, and pose manipulation, among others. With a deep understanding of complex linguistic structures, the model is able to produce visually compelling and semantically accurate outputs.

Qwen-Image is an ideal foundation model for developing creative visual content, paving the way for developers in building next-generation creative and AI-driven applications.

[Prompt: A movie poster. The first row is the movie title, which reads “Imagination Unleashed”. The second row is the movie subtitle, which reads “Enter a world beyond your imagination”. The third row reads “Cast: Qwen-Image”. The fourth row reads “Director: The Collective Imagination of Humanity”. The central visual features a sleek, futuristic computer from which radiant colors, whimsical creatures, and dynamic, swirling patterns explosively emerge, filling the composition with energy, motion, and surreal creativity. The background transitions from dark, cosmic tones into a luminous, dreamlike expanse, evoking a digital fantasy realm. At the bottom edge, the text “Launching in the Cloud, August 2025” appears in bold, modern sans-serif font with a glowing, slightly transparent effect, evoking a high-tech, cinematic aesthetic. The overall style blends sci-fi surrealism with graphic design flair—sharp contrasts, vivid color grading, and layered visual depth—reminiscent of visionary concept art and digital matte painting, 32K resolution, ultra-detailed.]

This article was originally published on Alizila written by Crystal Liu.

Community

Introducing Qwen-Image: Novel Model in Image Generation and Editing

Read previous post:

Read next post:

Alibaba Cloud Community

You may also like

Comments

Alibaba Cloud Community

Related Products

Tongyi Qianwen (Qwen)

Alibaba Cloud for Generative AI

AI Acceleration Solution

Platform For AI