Wan 2.5 Preview, regenerate images according to your intent
Based on an original image input, Wan 2.5 Preview regenerates the image on the basis of recognizing image content and spatial structure, while keeping the main subject position, composition relationships, and key elements consistent, so that the result better matches your expectations in style, layering, and visual expression. Whether replacing the scene environment, unifying the image style, or optimizing local details, Wan 2.5 Preview can complete adjustments without disrupting the original visual logic.
Log in to view your work
After you create an account, your images, videos, and creation history are saved so you can view, manage, and keep creating anytime.
Sign up free and start saving your creative history
Alibaba Cloud Tongyi Wanxiang Qwen Image｜ Speaking commercial visuals, one-shot generation
One paragraph of text generates poster-grade images with clear, readable text—especially strong in Chinese. Signs, posters, and packaging text look clearer and more realistic.
What can it do?
6 core capabilities for your creative needs
Text rendering
Signs, posters, and packaging text are clearer and more realistic, especially in Chinese
Precise editing
Change text, add or remove objects, switch styles, and enhance details—all reliably
Multi-image fusion
Multi-image input and output; combine A's clothes + B's pose + C's background
Style transfer
Turn photos into illustrations, posters, and various artistic styles
Commercial output
Posters, banners, product images—high-resolution output for commercial use
Complex understanding
Understands complex logic and accurately restores scenes described in prompts
Get started in 30 seconds
3 steps, easy enough for beginners
Choose mode
Text-to-image / Image editing
Write one sentence
Describe what you want
Generate and refine
Adjust text/style/composition until satisfied
Try these prompts:
Generate a coffee shop promo poster, title "Limited half price", subtitle "Today's special"
Generate a Bluetooth headset main image with "¥299" price tag
Generate a hotpot restaurant signboard containing "老北京" Chinese text
Case showcase Real results—see what AI can do for you
Generate a coffee shop promo poster, title "Limited half price", subtitle "Today's special"
Two modes, clear choices
Choose the right creation mode for your needs
Text-to-image mode
Image editing mode
Use "explicit instructions" when editing—what to change, what to change it to, keep the rest unchanged.
Use cases
8 scenarios—find the one for you
Event posters
Promo posters with Chinese title/price/time
Generate a 618 event poster, title "618狂欢购", subtitle "满200减50", bottom event date "6月1日-6月18日"
E-commerce main image
Edit text/price/promo
Change the price from "¥199" to "¥299" on this clothing image, keep the rest unchanged
Restaurant menu
Signature dishes and prices
Generate a Chinese restaurant menu, dish "招牌烤鸭", price "¥128", include dish image placeholder
App cover
Course/app series templates
Generate online course cover, title "Python入门", subtitle "零基础学编程", minimal modern style
Social media cover
Same theme, multiple styles
Generate Xiaohongshu cover, theme "秋季穿搭", 3 styles: vintage/fashion/minimal
Product packaging
Concept art with typography
Generate cosmetic gift box concept, logo "Rose", include "限量版" text
Image restoration
Detail enhancement and quality boost
Enhance this photo's quality—clearer details, more vibrant colors
Multi-image fusion
Person + clothing + pose combination
Apply figure 1's pose to figure 2's outfit, keep facial features unchanged
Why use Qwen Image on Dovoo AI?
Platform advantages for simpler creation
2s Faster speed
2-second generation, major efficiency boost
4K HD output
4K resolution support for commercial use
0 baseline Zero barrier
One sentence to generate, no design skills needed
FAQ
Best for commercial images with text: posters, banners, e-commerce main images, signboards, packaging. Especially strong at Chinese text rendering—clear, readable, consistent style.
Generate and edit on the same page
Start creating