Qwen Image

qwen-image

Qwen Image

qwen-image

Latest release (v2509, 25-09). Qwen-Image is an image generation foundation model that excels at complex text rendering and precise image editing, with support for multi-image editing.

Added Aug 7, 2025

Approx. Price

$0.020 per image

Model Type

both

Settings

Generation controls available for this model.

Images Per Run

Up to 4

Output images

Input Images

Up to 3

Reference/edit images accepted • Route max 30 MB

Output Sizes

Auto
Square
Portrait
Landscape
up to 1,024px

Enable Safety Checker

Default

Yes

Guidance Scale

Number

Default

2.5

How closely to follow the prompt (1-20)

Inference Steps

Number

Default

30

Number of inference steps (2-50)

Negative Prompt

Text

Default

N/A

Describe what you don't want in the image

Number of Images

Default

1

Resolution

Default

auto

Options (7)

auto (Auto (match input image for edits)), 1024x1024 (Square HD), 512x512 (Square), 768x1024 (Portrait (3:4)) +3 more

Seed

Number

Default

N/A

Random seed to control the stochasticity of image generation.

Benchmarks

Human preference benchmarks sourced from Artificial Analysis.

Text to Image

#91 / 147

ELO

1080.0

Appearances

4,812

95% CI

-8/8

Release Date 2025-08 · Matched as Qwen Image

Artificial Analysis API

Examples

Loading examples…