Qwen-Image-2.0-Pro-2026-03-03

Model IntroductionImage

Qwen-Image-2.1, a unified text-to-image generation and image editing model in the Qwen family.

Introduction

We are excited to open-source Qwen-Image-2.1, a unified text-to-image generation and image editing model in the Qwen family. With just 7B parameters in its visual generation component (32 Single-Stream DiT layers), Qwen-Image-2.1 balances generation quality, inference efficiency, and versatility.

Key Features

  • Compact and Efficient — A lightweight architecture with mixed-granularity attention and prefix KV cache reuse delivers strong image quality at low computational cost.
  • Native Transparency, Unified Creation and Editing — Generate regular or transparent (RGBA) images from text, edit transparent layers, and extract subjects from photographs—all in one model.
  • Versatile Editing — Support up to 10 reference images, specify local edits via circles, painted annotations, or separate masks, and preserve identity for people and products.
  • Realistic Textures and Refined Aesthetics — Improved typography, portrait lighting, and fine details for more visually compelling results.

Benchmarks

Benchmark (Area)Score*
Text-to-Image Overall (Qwen-Image-Bench) (Overall)57.84 (Eval)
Generation Quality (Quality + Aesthetics + Alignment) (Overall)57.45 (Eval)
Real-world Fidelity (Overall)51.83 (Eval)
Creative Generation (Overall)64.94 (Eval)
Prompt Adherence (Arena Skill) (Overall)22.2 (Eval)
Aesthetics (Arena Skill) (Overall)20.5 (Eval)
Text Rendering (Arena Skill) (Overall)20.0 (Eval)
Photorealism (Arena Skill) (Overall)18.1 (Eval)
Product, Branding & Commercial (Overall)19.5 (Eval)
Lumenfall Arena Elo (Overall)1225 (Eval)
SuperCLUE-Image Overall (Overall)75.68 (Eval)
SuperCLUE-Image Quality (Overall)87.25 (Eval)
SuperCLUE-Image Text-Image Consistency (Overall)46.67 (Eval)
SuperCLUE-Image Chinese Character Generation (Overall)84.52 (Eval)
SuperCLUE-Image Real-world Reproduction (Overall)83.02 (Eval)
SuperCLUE-Image Creation & Reasoning (Overall)76.93 (Eval)

API request example

Endpoint /v1/images/generations. Replace <API_KEY> with a key from the console.

`size` options: `1K` / `2K`. Allowed size: 1K / 2K.

curl https://api.ravenrouter.ai/v1/images/generations \
  -H "Content-Type: application/json" \
  -H "Authorization: Bearer <API_KEY>" \
  -d '{
  "model": "Qwen-Image-2.0-Pro-2026-03-03",
  "prompt": "a red apple on a wooden table",
  "size": "1K",
  "n": 1
}'

More models: API Reference → Examples by model

Qwen-Image-2.0-Pro-2026-03-03 API — $0.09/image | RavenRouter AI