aiaimodel.cards

Qwen-Image-3.0

Image modelCard summarized

Qwen-Image-3.0-Pro isn't just pursuing "good looks"—it's pursuing "usefulness"

Source: qwen-image-3.0-pro model page (Alibaba Cloud Model Studio)

Use it via API

$0.04 / image

Where to call it

  • Alibaba Cloud Model Studioqwen-image-3.0-pro

Supports

  • Image input
Full feature support

Good for

  • Infographics, posters, UI mock-ups and other text-heavy layouts
  • Small-text and multilingual typography
  • Photographic detail (micro-expressions, pores, hair)

Overview

Alibaba · 02Qwen-Image, front

Qwen-Image 3.0

Qwen-Image-3.0-Pro isn't just pursuing "good looks"—it's pursuing "usefulness"

Typical price
$0.04 / image
qwen-image-3.0-pro, 1K output (≤ 2.25 MP)
Max resolution
2048×2048 (total pixels 512×512 to 2048×2048)
Aspect ratios
—
Modes
text-to-image, image editing (1–3 input images)
Released
20 Jul 2026

Qwen-Image-3.0 is a closed text-to-image and image-editing model aimed at useful, information-dense images: prompts up to about 4,500 tokens, layouts such as newspapers, storyboards, menus and exam papers in one pass, text as small as 10 px in 12 languages, and realistic simulation of web pages, games and live streams. The Pro model favours quality; qwen-image-3.0 is faster and cheaper. model page; release notes

qwen-image-3.0-pro
Typical price
$0.04 / image
Max resolution
2048×2048 (total pixels 512×512 to 2048×2048)
Languages
Prompts in Chinese and English; native text rendering in 12 languages
Weights
Closed
Released
20 Jul 2026
  • text-to-image
  • image editing (1–3 input images)
  • multi-image fusion
  • dense layouts (newspapers, storyboards, menus, exam papers)
  • text rendering down to 10 px in 12 languages
  • prompt rewriting (direct or agent mode)
  • thinking mode for higher quality