Qwen Image Editing

curl --request POST \
  --url https://api.mixroute.ai/v1/images/generations

<?php

$curl = curl_init();

curl_setopt_array($curl, [
  CURLOPT_URL => "https://api.mixroute.ai/v1/images/generations",
  CURLOPT_RETURNTRANSFER => true,
  CURLOPT_ENCODING => "",
  CURLOPT_MAXREDIRS => 10,
  CURLOPT_TIMEOUT => 30,
  CURLOPT_HTTP_VERSION => CURL_HTTP_VERSION_1_1,
  CURLOPT_CUSTOMREQUEST => "POST",
]);

$response = curl_exec($curl);
$err = curl_error($curl);

curl_close($curl);

if ($err) {
  echo "cURL Error #:" . $err;
} else {
  echo $response;
}

package main

import (
	"fmt"
	"net/http"
	"io"
)

func main() {

	url := "https://api.mixroute.ai/v1/images/generations"

	req, _ := http.NewRequest("POST", url, nil)

	res, _ := http.DefaultClient.Do(req)

	defer res.Body.Close()
	body, _ := io.ReadAll(res.Body)

	fmt.Println(string(body))

}

require 'uri'
require 'net/http'

url = URI("https://api.mixroute.ai/v1/images/generations")

http = Net::HTTP.new(url.host, url.port)
http.use_ssl = true

request = Net::HTTP::Post.new(url)

response = http.request(request)
puts response.read_body

Qwen Image Editing

curl --request POST \
  --url https://api.mixroute.ai/v1/images/generations

<?php

$curl = curl_init();

curl_setopt_array($curl, [
  CURLOPT_URL => "https://api.mixroute.ai/v1/images/generations",
  CURLOPT_RETURNTRANSFER => true,
  CURLOPT_ENCODING => "",
  CURLOPT_MAXREDIRS => 10,
  CURLOPT_TIMEOUT => 30,
  CURLOPT_HTTP_VERSION => CURL_HTTP_VERSION_1_1,
  CURLOPT_CUSTOMREQUEST => "POST",
]);

$response = curl_exec($curl);
$err = curl_error($curl);

curl_close($curl);

if ($err) {
  echo "cURL Error #:" . $err;
} else {
  echo $response;
}

package main

import (
	"fmt"
	"net/http"
	"io"
)

func main() {

	url := "https://api.mixroute.ai/v1/images/generations"

	req, _ := http.NewRequest("POST", url, nil)

	res, _ := http.DefaultClient.Do(req)

	defer res.Body.Close()
	body, _ := io.ReadAll(res.Body)

	fmt.Println(string(body))

}

require 'uri'
require 'net/http'

url = URI("https://api.mixroute.ai/v1/images/generations")

http = Net::HTTP.new(url.host, url.port)
http.use_ssl = true

request = Net::HTTP::Post.new(url)

response = http.request(request)
puts response.read_body

Edit images using Qwen Image models with the native synchronous Alibaba request structure. POST https://api.mixroute.ai/v1/images/generations

input and parameters are native Qwen fields. Put the prompt in input.messages[].content[].text, not in an additional top-level prompt. This page documents synchronous requests, not the separate asynchronous image-synthesis API.

Models

Model ID	Capabilities and Limits
`qwen-image-2.0-pro`	Generation and editing; up to six outputs; 1300-token prompt limit.
`qwen-image-2.0`	Generation and editing; up to six outputs; 1300-token prompt limit.
`qwen-image-edit-max`	Editing; up to six outputs; custom output dimensions.
`qwen-image-edit-plus`	Editing; up to six outputs; custom output dimensions.
`qwen-image-edit`	Editing; one output; no custom size or prompt_extend setting.

Request Parameters

Field	Type	Required	Description
`model`	string	Yes	Complete model ID from the model table.
`input`	object	Yes	Native input object.
`input.messages`	object[]	Yes	Exactly one message; this interface is single-turn.
`input.messages[].role`	string	Yes	Must be `user`.
`input.messages[].content`	object[]	Yes	One text item plus 1-3 image items, in reference order.
`input.messages[].content[].text`	string	Yes	Positive prompt. Qwen Image 2.0: at most 1300 tokens; other models: 800 tokens. Excess input is truncated. Do not include multiple text items.
`input.messages[].content[].image`	string	Yes	Public HTTP/HTTPS URL, accessible native temporary OSS URL, or Base64 data URI. Supply 1-3 images. Output proportions default to the last source image.
`parameters`	object	No	Native output and prompt-processing settings.
`parameters.n`	integer	No	Default 1. Qwen Image 2.0 and Edit Max/Plus allow 1-6; qwen-image-edit only allows 1.
`parameters.size`	string	No	Output size formatted as `width*height`. See the model/mode limits below; it is not the OpenAI `widthxheight` format.
`parameters.negative_prompt`	string	No	Content to exclude. Supports Chinese/English; at most 500 characters, with excess text truncated.
`parameters.prompt_extend`	boolean	No	Default true; rewrites the positive prompt without changing the negative prompt. Not supported by qwen-image-edit.
`parameters.watermark`	boolean	No	Default false; true adds the Qwen-Image watermark in the lower-right corner.
`parameters.seed`	integer	No	Range 0-2147483647. Omit for a random seed. Reusing a seed can improve consistency but does not guarantee identical output.

Size Rules

Model	Output Size
Qwen Image 2.0 / Pro	Total pixels 262144-4194304. Without size, about `1024*1024` total pixels and the last input image aspect ratio.
Qwen Image Edit Max / Plus	Each edge 512-2048 px. Without size, about `1024*1024` total pixels and the last input image aspect ratio.
Qwen Image Edit	Do not send size. The model chooses dimensions around `1024*1024` total pixels, following the input aspect ratio.

Requested dimensions may be aligned to nearby multiples of 16. Input formats: JPG/JPEG/PNG/BMP/TIFF/WebP/GIF, at most 10 MB per image; only the first GIF frame is processed. Dimensions of 384-3072 px per edge are recommended, not a hard minimum/maximum. Native OSS references must be accessible to the upstream account.

Examples

Use your MixRoute key in the MIXROUTE_API_KEY environment variable. Requests use Authorization: Bearer .... Replace source-image placeholders with accessible images or local files before editing.

curl --request POST "https://api.mixroute.ai/v1/images/generations" \
  --header "Authorization: Bearer $MIXROUTE_API_KEY" \
  --header "Content-Type: application/json" \
  --data '{
  "model": "qwen-image-2.0-pro",
  "input": {
    "messages": [
      {
        "role": "user",
        "content": [
          {
            "image": "https://example.com/source.png"
          },
          {
            "text": "Make the mug blue while preserving its shape, lighting, and background."
          }
        ]
      }
    ]
  },
  "parameters": {
    "size": "1024*1024",
    "n": 1,
    "prompt_extend": true,
    "watermark": false
  }
}'

Python

import json
import os
import requests

payload = json.loads(r'''
{
  "model": "qwen-image-2.0-pro",
  "input": {
    "messages": [
      {
        "role": "user",
        "content": [
          {
            "image": "https://example.com/source.png"
          },
          {
            "text": "Make the mug blue while preserving its shape, lighting, and background."
          }
        ]
      }
    ]
  },
  "parameters": {
    "size": "1024*1024",
    "n": 1,
    "prompt_extend": true,
    "watermark": false
  }
}
''')
response = requests.post(
    "https://api.mixroute.ai/v1/images/generations",
    headers={"Authorization": "Bearer " + os.environ["MIXROUTE_API_KEY"]},
    json=payload,
    timeout=180,
)
response.raise_for_status()
result = response.json()
for choice in result.get("output", {}).get("choices", []):
    for part in choice.get("message", {}).get("content", []):
        if "image" in part:
            print(part["image"])

Response

Read image URLs from output.choices[].message.content[].image. Outputs are PNG; hosted links expire after 24 hours. request_id is for troubleshooting, not a video-task polling ID. This is not the OpenAI data[].b64_json response structure. Qwen image generation

Popular Models

OpenAI SDK Usage

System API

Text Series

Embedding Series

Image Series

Video Series

Others

Qwen Image Editing

Models

Request Parameters

Size Rules

Examples

Python

Response

​Models

​Request Parameters

​Size Rules

​Examples

​Python

​Response

Models

Request Parameters

Size Rules

Examples

Python

Response