Synchronous and Asynchronous Calls for Image Generation and Editing 2.6
/api/v1/image/generationHeader Parameters
Use Bearer Token authentication. Format: Authorization: Bearer sk-xxxxxx
Used to enable streaming output; this field must be set to enable only when parameters.enable_interleave=true.
Asynchronous processing configuration parameter, must be set to enable.
Request Body
application/json
Model name. Fixed value: wan2.6-image
"wan2.6-image"Basic input information.
Request content array. Currently, only single-turn conversations are supported, i.e., passing a single set of role and content parameters, while multi-turn conversations are not supported.
The role of the message. This parameter is always set to user.
Message content array.
Positive prompts are used to describe the desired content, style, and composition of the generated image. Both Chinese and English are supported, with a maximum length of 2000 characters. Each Chinese character, letter, number, or symbol counts as one character, and any excess will be automatically truncated. Example value: Generate scrambled eggs with tomatoes referencing the style of this image. Note: The content array must contain one and only one object with a text field.
Enter the URL or Base64-encoded string of the input image. Image constraints: Image formats: JPEG, JPG, PNG (alpha channel not supported), BMP, and WEBP. Image resolution: Both width and height must be within the range of [240, 8000] pixels. File size: Up to 10MB. Image count limit: The number of input images depends on the parameters.enable_interleave parameter. When enable_interleave=true (interleaved text and image output), 0 to 1 image can be input. When enable_interleave=false (image editing), 1 to 4 images must be input. When inputting multiple images, pass multiple image objects in the content array and define the image sequence according to the array order.
Image processing parameters.
Negative prompt, used to describe unwanted elements in the image and constrain the visual output. Supports both Chinese and English, with a maximum length of 500 characters; any excess will be automatically truncated.
Output image resolution parameter, supporting two methods: referencing the input image aspect ratio and direct specification. When enable_interleave=false (i.e., image editing mode): Method 1: Reference input image aspect ratio (recommended). Available output resolution tiers: 1K (default) and 2K. 1K: Total output pixels close to 12801280, with the aspect ratio matching the last input image. 2K: Total output pixels close to 20482048, with the aspect ratio matching the last input image. Method 2: Specify the width and height pixel values of the generated image. Total pixels within [768768, 20482048], and aspect ratio within [1:4, 4:1]. The actual output image pixel values are multiples of 16 closest to the specified values. When enable_interleave=true (i.e., interleaved text and image output mode): Method 1: Reference input image aspect ratio (default). If the total input image pixels ≤ 12801280, the total output pixels and aspect ratio match the input image. If the total input image pixels > 12801280, the total output pixels are close to 12801280, with the aspect ratio matching the input image. Method 2: Specify the width and height pixel values of the generated image. Total pixels within [768768, 12801280], and aspect ratio within [1:4, 4:1]. The actual output image pixel values are multiples of 16 closest to the specified values. Recommended resolutions for common aspect ratios: 1:1: 12801280, 2:3: 8001200, 3:2: 1200800, 3:4: 9601280, 4:3: 1280960, 9:16: 7201280, 16:9: 1280720, 21:9: 1344*576.
Control image generation mode: false: Default value, indicating image editing mode (supports multi-image input and subject-consistent generation). Usage: Editing, style transfer, or subject-consistent generation based on 1 to 4 input images. Input: At least one reference image must be provided. Output: Can generate 1 to 4 resulting images. true: Indicates enabling the mixed text-and-image output mode (supports passing only one image or no image). Usage: Generating content with both text and images based on text descriptions, or text-to-image generation. Input: No image required (text-to-image), or up to 1 reference image can be provided. Output: Generates mixed content containing text and images.
Specifies the number of images to generate. The value range and meaning of this parameter depend on the state of enable_interleave (mode switch): When enable_interleave=false (image editing mode): Function: Directly controls the number of generated images. Value range: 1 to 4, with a default value of 4. It is recommended to set this value to 1 during the testing phase to verify the effects at a low cost. When enable_interleave=true (text-image interleaved mode): Restriction: This parameter defaults to 1 and must be fixed at 1. If set to any other value, the API will return an error. Note: In this mode, to control the maximum number of generated images, please use the max_images parameter.
Takes effect only in text-image interleaved mode (i.e., enable_interleave=true). Purpose: Specifies the maximum number of images the model can generate in a single response. Value range: 1 to 5, default is 5. Note: This parameter only represents the upper limit. The actual number of images generated is determined by model inference and may be less than the specified value (e.g., if set to 5, the model may generate only 3 images depending on the content).
Takes effect only in image editing mode (i.e., enable_interleave = false). Whether to enable the intelligent Prompt rewriting feature. This feature only optimizes and polishes positive prompts without altering negative prompts. true: Default value, enables intelligent rewriting. false: Disables intelligent rewriting and uses the original prompts.
Controls whether the response is streamed. In image interleaving mode (i.e., enable_interleave = true), it must be set to true. false: Default value, non-streaming output. true: Streaming output.
Whether to add a watermark. The watermark is located in the bottom right corner of the image with the fixed text "AI-generated". false: Default value, no watermark is added. true: Add a watermark.
Random seed, value range [0, 2147483647]. Using the same seed parameter value keeps the generated content relatively stable. If not provided, the algorithm will automatically use a random seed. Note: The model generation process is probabilistic, and even with the same seed, completely identical results cannot be guaranteed every time.
Response Body
application/json
Success
TypeScript Definitions
Use the response body type in TypeScript.
Status Code
Data Object
Result Set
Usage Information
Number of Images
Video Duration in Seconds
Input Token Count
Total Tokens
Output Token Count
Validity Period
Task ID
Task Status
Status Information
Request ID
curl -X POST "https://api.ygw.ai/api/v1/image/generation" \ -H "Authorization: string" \ -H "Content-Type: application/json" \ -d '{ "model": "wan2.6-image", "input": { "messages": [ { "role": "string", "content": [ { "text": "string" } ] } ] } }'{
"code": 0,
"data": {
"urls": [
"string"
],
"usage": {
"images": 0,
"seconds": 0,
"input_tokens": 0,
"total_tokens": 0,
"output_tokens": 0
},
"expired": "string",
"task_id": "string",
"task_status": "string"
},
"message": "string",
"request_id": "string"
}