qwen-image-2512

Qwen‑Image‑2512 is a text-to-image diffusion model in the Qwen family. With about 20 billion parameters, it generates images from Chinese and English prompts. It improves human realism, reduces the typical artificial look of generated images and accurately renders text inside images. It supports styles ranging from photorealistic scenes to illustration for professional generation.

• PROVIDER LOCATION: Non-EU
• PARAMETERS: ~20B
• MAX CONTEXT: N/A
• LICENSE: Apache 2.0
• FEATURES: Image generation from text prompts
• OUTPUT FORMATS: Image (PNG/JPEG)

• Photorealistic image generation
• Visual content creation for marketing
• Illustrations and concept art
• Mockup and visual prototype generation

curl --request POST \
--url https://apiaruba.it/aiplatform/v1/images/generations \
--header 'content-type: application/json' \
--header 'x-aruba-api-key: *****' \
--data '{
"prompt": "genera un gatto seduto su un davanzale al tramonto",
"model": "Qwen-Image-2512"
}'

{
"created": 1787586338,
"background": null,
"data": [
{
"b64_json": "*****",
"revised_prompt": null,
"url": null
}
],
"output_format": "png",
"quality": null,
"size": "512x512",
"usage": {
"total_tokens": 0,
"input_tokens": 0,
"input_tokens_details": {
"image_tokens": 0,
"text_tokens": 0
},
"output_tokens": 0,
"output_tokens_details": null
}
}

• Qwen‑Image‑Edit‑2511

• Hugging Face
• Blog Qwen
• Demo