Skip to main content
Alibaba QwenAug 5, 2026

Qwen Image 3

qwen-image-3AnonymizedUncensored

Qwen Image 3 is the standard tier of Alibaba’s Qwen Image 3.0, offering cost-efficient text-to-image generation at 1K or 2K with strong text rendering. It pairs with Qwen Image 3 Edit.

$0.04
per image
2K
max resolution, 2 sizes
8
aspect ratios, 1:1 default

Same prompt, other models

Every image model renders the same reference prompts, so their results line up side by side.

Photorealistic close-up portrait of a weathered fisherman at golden hour, 85mm lens, shallow depth of field, natural skin texture

Pricing

Billed per image in USD, by resolution. Each resolution is drawn to scale.

1K · default2K
Per image$0.04$0.04
Edit$0.04
Extra input image$0.0034
2x upscale$0.02
4x upscale$0.08

Estimate a batch

Resolution
$0.04per image
$40.00for 1,000 images

Against other image models

How it comparesamong 51 image models
Price per imagecheaper than 64%$0.04
Prompt limitlonger than 42%10,000 characters

Further right is better. Hover a dot for its model.

Capabilities

What the family does, and every shape it renders.

Generateqwen-image-3
Editqwen-image-3-edit · combines images
Upscale2x $0.02 · 4x $0.08
Repeatableseed

Aspect ratios

1:13:216:921:99:162:33:44:5

Specs

Model ID
qwen-image-3
Edit model ID
qwen-image-3-edit
Endpoint
POST /api/v1/image/generate
Resolutions
1K, 2K · default 1K
Aspect ratios
8 · default 1:1
Prompt limit
10,000 characters
Steps
default 20, max 50
Privacy
Anonymized
Weights
Closed
Added
Aug 5, 2026

Other versions

  1. Qwen ImageJan 2025
  2. Qwen Image 2Mar 2026
  3. Qwen Image 3Aug 2026

Quickstart

Generate with Venice’s image endpoint, or with the OpenAI-compatible one from any OpenAI SDK.

curl https://api.venice.ai/api/v1/image/generate \
-H "Authorization: Bearer $VENICE_API_KEY" \
-H "Content-Type: application/json" \
-d '{
"model": "qwen-image-3",
"prompt": "A retro travel poster with the bold headline \"VENICE\" in large condensed serif type, sunset color palette",
"resolution": "1K",
"aspect_ratio": "1:1"
}'
Base URL https://api.venice.ai/api/v1

Parameters

ParameterValuesNotes
modelqwen-image-3Required
promptUp to 10,000 charactersRequired
resolution1K | 2KDefault 1K
aspect_ratio8 ratiosDefault 1:1
stepsUp to 50Default 20
seedIntegerRepeats a result
variants1 – 4Images per request
formatwebp | png | jpegDefault webp
safe_modetrue | falseDefault true, which blurs adult content