Wan2.7 Image / Pro

Generate, edit, and create image sequences with wan2.7-image and wan2.7-image-pro through AnyFast.

Wan2.7 Image is Alibaba’s image generation and editing model family. Through AnyFast, both variants use the same synchronous endpoint and DashScope-compatible message structure.

Model ID	Description
wan2.7-image-pro	Professional variant. Text-to-image requests without sequential output support up to 4K.
wan2.7-image	Standard variant optimized for faster generation. Supports up to 2K.

Capabilities

• Text-to-image — Generate one or more images from a text prompt
• Image editing — Edit up to 9 input images with text instructions
• Interactive editing — Limit edits to regions defined by bbox_list
• Sequential generation — Create a related image sequence from text or image references
• Multi-image reference — Place multiple { "image": "..." } items in one content array

parameters.prompt_extend defaults to true. For special-content requests, select the Special-Ns resource group and set this parameter to false.

The endpoint accepts the existing flat OpenAI-compatible format and the newly adapted nested DashScope-compatible format shown in the AnyFast examples. Use the flat format for simple generation or single-image editing, and the nested format for multiple images, region editing, or sequential generation.

Request format

Send requests to POST /v1/images/generations. The following examples show both supported request formats.

Text-to-image

curl https://www.anyfast.ai/v1/images/generations \
  -H "Authorization: Bearer YOUR_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "model": "wan2.7-image",
    "prompt": "A cinematic flower shop with detailed windows and a wooden door",
    "size": "2K",
    "n": 1,
    "watermark": false,
    "thinking_mode": true
  }'

curl https://www.anyfast.ai/v1/images/generations \
  -H "Authorization: Bearer YOUR_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "model": "wan2.7-image-pro",
    "prompt": "Replace the vase with a bouquet of red roses",
    "image": "https://example.com/room.png",
    "size": "2K",
    "n": 1
  }'

curl https://www.anyfast.ai/v1/images/generations \
  -H "Authorization: Bearer YOUR_API_KEY" \
  -H 'X-DashScope-DataInspection: {"input":"disable","output":"disable"}' \
  -H "Content-Type: application/json" \
  -d '{
    "model": "wan2.7-image-pro",
    "input": {
      "messages": [
        {
          "role": "user",
          "content": [
            {"image": "https://example.com/person.png"},
            {"image": "https://example.com/background.png"},
            {"text": "Place the person from image 1 into the setting from image 2"}
          ]
        }
      ]
    },
    "parameters": {
      "size": "2K",
      "n": 1,
      "watermark": false,
      "thinking_mode": true,
      "prompt_extend": true
    }
  }'

curl https://www.anyfast.ai/v1/images/generations \
  -H "Authorization: Bearer YOUR_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "model": "wan2.7-image-pro",
    "input": {
      "messages": [
        {
          "role": "user",
          "content": [
            {"image": "https://example.com/room.png"},
            {"text": "Replace the selected vase with red roses"}
          ]
        }
      ]
    },
    "parameters": {
      "size": "2K",
      "n": 1,
      "bbox_list": [
        [[420, 280, 760, 720]]
      ],
      "prompt_extend": true
    }
  }'

curl https://www.anyfast.ai/v1/images/generations \
  -H "Authorization: Bearer YOUR_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "model": "wan2.7-image-pro",
    "input": {
      "messages": [
        {
          "role": "user",
          "content": [
            {"text": "Create a four-image cinematic sequence of the same orange cat across spring, summer, autumn, and winter"}
          ]
        }
      ]
    },
    "parameters": {
      "size": "2K",
      "enable_sequential": true,
      "n": 4,
      "watermark": false,
      "prompt_extend": true
    }
  }'

Input rules

Flat OpenAI-compatible format

Field	Type	Description
model	string	wan2.7-image or wan2.7-image-pro
prompt	string	Chinese or English prompt, up to 5,000 characters
image	string	Optional public image URL or Base64 data URI for single-image editing
size	string	Output size such as 1K or 2K; Pro text-to-image also supports 4K
n	integer	Number of generated images
thinking_mode	boolean	Enables thinking mode for text-to-image
watermark	boolean	Adds the fixed AI-generated watermark
seed	integer	Random seed from 0 to 2,147,483,647

Nested DashScope-compatible format

Field	Requirement
input.messages	One message with role: "user"
content[].text	Chinese or English prompt, up to 5,000 characters
content[].image	Public HTTP/HTTPS URL or Base64 data URI
Image formats	JPEG, JPG, PNG without transparency, BMP, WEBP
Image size	Up to 20 MB per image; width and height 240–8,000 px
Image aspect ratio	From 1:8 to 8:1
Image count	0–9 images; array order defines image order

Parameters

Parameter	Type	Default	Description
parameters.size	string	2K	wan2.7-image-pro: 1K, 2K, or 4K for non-sequential text-to-image; other modes support up to 2K. wan2.7-image: 1K or 2K.
parameters.n	integer	1 or 12	Non-sequential mode: 1–4, default 1. Sequential mode: maximum output count 1–12, default 12. Billing follows successfully generated images.
parameters.enable_sequential	boolean	false	Enables text-to-sequence or image-to-sequence generation.
parameters.thinking_mode	boolean	true	Improves text-to-image quality. Only applies with no image input and sequential mode disabled.
parameters.prompt_extend	boolean	true	Controls AnyFast prompt extension. For special-content requests, select the Special-Ns resource group and set this parameter to false.
parameters.bbox_list	array	—	Region boxes for interactive editing. Provide one list per input image; each image supports up to two [x1, y1, x2, y2] boxes.
parameters.color_palette	array	—	3–10 HEX colors and percentage ratios totaling 100.00%; available only when sequential mode is disabled.
parameters.watermark	boolean	false	Adds the fixed AI-generated watermark in the lower-right corner.
parameters.seed	integer	Random	Integer from 0 to 2,147,483,647. Equal seeds can produce similar, but not identical, results.

When images are supplied, the output uses the aspect ratio of the last input image and scales it to the selected resolution. Generated dimensions can differ slightly from custom dimensions.

Response

AnyFast returns the generated image in data and preserves the upstream Wan response under metadata.

{
  "data": [
    {
      "url": "https://dashscope-result.example.com/generated.png",
      "b64_json": "",
      "revised_prompt": ""
    }
  ],
  "created": 1784274013,
  "metadata": {
    "output": {
      "choices": [
        {
          "finish_reason": "stop",
          "message": {
            "content": [
              {
                "image": "https://dashscope-result.example.com/generated.png",
                "type": "image"
              }
            ],
            "role": "assistant"
          }
        }
      ],
      "finished": true
    },
    "usage": {
      "image_count": 1,
      "input_tokens": 25,
      "output_tokens": 2,
      "size": "2048*2048",
      "total_tokens": 27
    },
    "request_id": "f30f2d2b-e0ee-9a47-ba5e-68b66360fe47"
  }
}

Generated URLs are retained for 24 hours by the upstream service. Download outputs promptly rather than storing these URLs permanently.

API Reference

Open the Wan2.7 Image / Pro API reference.