{
"created": 123,
"data": [
{
"url": "<string>",
"b64_json": "<string>",
"revised_prompt": "<string>"
}
],
"metadata": {}
}Call Ali qwen-image-2.0-pro for single-image editing or multi-image fusion.
The public contract keeps top-level model, input, and response_format, while exposing n, size, watermark, prompt_extend, negative_prompt, and seed as top-level fields. Input images are provided in input.messages[0].content[].image, and the edit instruction is the text item in the same content array. Top-level parameters is not accepted.
{
"created": 123,
"data": [
{
"url": "<string>",
"b64_json": "<string>",
"revised_prompt": "<string>"
}
],
"metadata": {}
}Bearer authentication header of the form Bearer <token>, where <token> is your auth token.
Image editing model name.
qwen-image-2.0-pro Show child attributes
Unified response format.
url, b64_json Number of output images. The gateway maps it to upstream parameters.n.
1 <= x <= 6Output image resolution in widthxheight or width*height format. The gateway normalizes it and maps it to upstream parameters.size.
Whether to add watermark. The gateway maps it to upstream parameters.watermark.
Whether to enable prompt rewriting. The gateway maps it to upstream parameters.prompt_extend.
Negative prompt. The gateway maps it to upstream parameters.negative_prompt.
Random seed. Explicit 0 is preserved and mapped to upstream parameters.seed.