Qwen Image 3

Qwen Image 3 (qwen/qwen-image-3) is Alibaba’s model for image generation and editing: it renders in-image text accurately — Cyrillic included — and holds detail down to 10 pixels. Two endpoints: POST /v1/images/generations for text-to-image and POST /v1/images/edits for working from a reference image. The model is paid — generations are charged from your balance.

Parameters

Field	Description
model	qwen/qwen-image-3 (or the short qwen-image-3)
prompt	What to draw, required
n	How many images, 1 to 4; each one is billed separately
aspect_ratio	1:1, 1:2, 1:4, 2:1, 2:3, 3:2, 3:4, 4:1, 4:3, 4:5, 5:4, 9:16, 16:9; defaults to 1:1
resolution	1k or 2k; defaults to 1k
size	Explicit WxH — an alternative to the aspect_ratio + resolution pair
quality	The model has no such knob: passing the parameter returns 400

Qwen Image 3 sizes are computed from the long side: 1024 pixels at 1k and 2048 at 2k. Common combinations:

Aspect	1k	2k
1:1	1024x1024	2048x2048
16:9	1024x576	2048x1152
9:16	576x1024	1152x2048
4:3	1024x768	2048x1536
4:1	1024x256	2048x512

The full list of 26 allowed size values is the sizes field in the catalog; a value outside the list returns 400. There is no 4k resolution — a request with it also returns 400.

Examples

Text-to-image

Python

import base64

img = client.images.generate(
    model="qwen/qwen-image-3",
    prompt="Street festival poster: bold typography, stencil style, two colors",
    extra_body={"aspect_ratio": "2:3", "resolution": "2k"},
)
with open("poster.png", "wb") as f:
    f.write(base64.b64decode(img.data[0].b64_json))

cURL

curl https://api.mixen.ai/v1/images/generations \
  -H "Authorization: Bearer $MIXEN_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "model": "qwen/qwen-image-3",
    "prompt": "Street festival poster: bold typography, stencil style, two colors",
    "aspect_ratio": "2:3",
    "resolution": "2k"
  }'

A vertical poster with an explicit size

{
  "model": "qwen/qwen-image-3",
  "prompt": "Running app cover: night city, rhythm, a neon route grid",
  "size": "1364x2048"
}

Editing from a reference

The model accepts reference images (references: true), so /v1/images/edits with multipart/form-data is available:

curl https://api.mixen.ai/v1/images/edits \
  -H "Authorization: Bearer $MIXEN_API_KEY" \
  -F model=qwen/qwen-image-3 \
  -F image=@menu.png \
  -F prompt="Change the price on the menu to 450, keep everything else untouched"

The answer always carries b64_json — files are not hosted. Both endpoints answer synchronously; set your HTTP client timeout to at least 120 seconds.

Price

4.0108 ₽ ($0.045) per image — the flat catalog price. Current values — catalog.

Strengths and limits

• Accurate in-image text — Cyrillic included — and fine detail down to 10 pixels.
• Thirteen frame formats: from square 1:1 to the elongated 4:1 and 1:4 strips.
• Predictable sizes: the long side is 1024 at 1k and 2048 at 2k.
• Editing from a reference via /v1/images/edits, up to 4 references.
• No 4k resolution and no quality knob — both parameters return 400.
• Balance only — the model is not part of the free quota.

All models — in the catalog. The general image workflow — images guide.