AI Agents

Updated time：2024-06-18

Overview

AI agent is a core concept in WeShop AI, representing different agents and tools we have that tailored to your needs. Please refer to https://www.weshop.ai/agent (opens in a new tab) if you need more information. In WeShop AI, users need to select the appropriate agent(s) to start the task. Each agent includes detailed instructions to help users precisely choose the right agent.

Relevant interfaces of AI agents, please refer to AI Agents Reference Interface，Below is an introduction to the generic interfaces for different AI agents. For detailed usage methods of different agents, please visit the specific agent for more information.

AI Model

Showcase clothing with a variety of models that aligns with your brand’s aesthetic

Basic Information

AgentName

aimodel

AgentVersion

- v1.0

Note

When executing the task, please select at least one or more parameters from fashionModelId, locationId, and textDescription to execute.
- If you prefer to use the template, you can provide the locationId and fashionModelId parameters. You can query the valid locationId and fashionModelId values through the AI entity's info API.
- If you want to describe the image you want to generate with words/phrases/sentences, you can assign a value to the textDescription parameter.
- Mask settings:
  - If you only need to keep the clothing on the model unchanged, please set maskType to "autoApparelSegment" when executing the task. The AI entity will use automatic apparel segmentation technology to generate the mask image, and you do not need to provide customMask and customMaskUrl parameters at this time.
  - If you only need to change the background without replacing the model's face, please set maskType to "autoSubjectSegment" when executing the task. The AI entity will use automatic subject segmentation technology to generate the mask image, and you do not need to provide customMask and customMaskUrl parameters at this time.
  - If you need to select the areas to keep unchanged on your own instead of using automatic segmentation, please prepare the mask image before the operation and set maskType to "custom" when executing the task. In this case, you need to provide either customMask or customMaskUrl parameter.

Tasks

Create Task

Request Parameters

initParams

Parameters Sample：

{
    "initParams": {
        "taskName": "Create Task Api Test",
        "originalImage": "https://ai-image.weshop.ai/xxxxxxxx.png"
    }
}

Parameter Name	Required	Type	Description	Notes
taskName	No	string	Task name	Do not exceed 20 characters
originalImage	Yes	string	Original image	When using an external image link, you must ensure that the image link is publicly accessible on the internet.

Response Parameters

data

Parameter Name	Required	Type	Description	Notes
taskId	Yes	string	Task ID - Unique Identifier	/

Task Execution

Request Parameters

params

Parameters Sample

// Automatic Apparel Segmentation
{
    "params": {
        "generatedContent": "referToOrigin",
        "maskType": "autoApparelSegment", 
        "locationId": 2093,
        "fashionModelId": 20053,
        "tagIds": [2006],
        "textDescription": "A young caucasian female, highly detailed skin, realistic detailed eyes, natural skin texture, confident expression, standing on the street, sunny day, bright, summer afternoon, sharp focus, high quality, taken with iPhone, hyper-realistic.",
        "negTextDescription": "dark clouds",
    }
}
 
// Automatic Subject Segmentation 
{
    "params": {
        "generatedContent": "freeCreation",
        "maskType": "autoSubjectSegment",
        "locationId": 2093
    }
}
 
// Upload Custom Mask
{
    "params": {
        "generatedContent": "freeCreation",
        "maskType": "custom",
        "customMask": "<mask image base64 encoded>",
        "textDescription": "A young caucasian female, highly detailed skin, realistic detailed eyes, natural skin texture, confident expression, standing on the street, sunny day, bright, summer afternoon, sharp focus, high quality, taken with iPhone, hyper-realistic.",
        "negTextDescription": "dark clouds",
    }
}

Parameter Name	Required	Type	Description	Notes
generatedContent	Yes	string	Select algorithm preference： "freeCreation"：Free creation "referToOrigin"：Refer to the original image	1. FreeCreation free creation will make the resulting image more free, without being constrained by the style of the uploaded original image. 2. referToOrigin reference original image will make the resulting image refer to the style of the uploaded original image during the generation process.
maskType	Yes	string	Select mask type: "autoApparelSegment": Automatic apparel segmentation "autoSubjectSegment": Automatic subject segmentation "custom": Manually upload mask information	1. autoApparelSegment automatic apparel segmentation is consistent with the automatic segmentation on the web page. 2. autoSubjectSegment automatic subject segmentation is for subject background segmentation, different from manual selection on the web page. 3. For manual upload, ensure the uploaded mask image has the same dimensions as the original image.
customMask	No	string	Upload the base64 encoded mask image	1. When the mask type is "custom", this parameter must be uploaded, otherwise no need to upload. 2. The area outside the mask region is transparent. 3. Use base64 encoding, without the prefix "data:image/png;base64". 4. Must have the same dimensions as the original image.
customMaskUrl	No	string	URL of the uploaded mask image	When using an external image link, you must ensure that the image link is publicly accessible on the internet. The image format requirements are the same as customMask.
locationId	No	int	Location	Optional
locationTagIds	No	array[int]	Location tags	Optional, please refer to Locations
fashionModelId	No	int	Model	Optional
fashionModelTagIds	No	array[int]	Model tags	Optional, please refer to FashionModels
textDescription	No	string	Text description of the content (optional)	Optional
negTextDescription	No	string	Description of content you don't want to appear (optional)	Optional
resultImageFormat	No	string	Select the format of the generated result image, options: "jpg" "png" (default is png)	Default is png
resizeToOriginalImage	No	bool	Whether to resize the generated result image to the original image dimensions	Default is false
batchCount	No	int	Set the number of images to generate in this execution, minimum is 1, maximum is 16	Default is 4, values out of range will use the default

Response Parameters

data

Parameter Name	Required	Type	Description	Notes
executionId	Yes	string	Execution ID - the unique identifier	/

Query Task

Request Parameters

json format, parameters sample

{
    "executionId":"<execution id>"
}

Parameter Name	Required	Type	Description	Notes
executionId	No	string	Execution ID	/
taskId	No	string	Task ID	If the taskId parameter is passed, at most the last 20 execution records will be returned.

Response Parameters

initParams

Current task's initial information

Parameter Name	Required	Type	Description	Notes
taskName	Yes	string	Task name	/
originalImage	Yes	string	Original image	/

executions

Execution information

Parameter Name	Required	Type	Description	Notes
executionId	Yes	string	Execution ID	/
status	Yes	string	Execution Status： Segmenting：In automatic segmentation Pending：Queuing Running ：In progress Success：Completed	/
params	Yes	object	Task execution related parameters	/
executionTime	Yes	string	Task execution timestamp	/
results	Yes	array[object]	Task execution result information	/

result

Execution result

Parameter Name	Required	Type	Description	Notes
status	Yes	string	Execution results Running Success Failed Pending	/
image	Yes	string	The generated final image	/
progress	No	string	Processing progress, e.g. 0.71 means 71% has been processed	/
error	No	string	Information if processing failed	When the status is "Failed", this field will definitely be returned.

Information Query

Query information related to the AI agent. By passing in the agent name (agentName) and version number (agentVersion), you can obtain the corresponding configuration parameter information.

Request Parameters

Parameter Name	Required	Type	Description	Notes
agentName	Yes	string	agent name： "aimodel"	/
agentVersion	Yes	string	agent version： "v1.0"	/

Response Parameters

data

Parameter Name	Required	Type	Description	Notes
locations	Yes	array[object]	Location information	1. You can obtain complete Locations information through the AI agent information query interface (apiReference#info) 2. Please refer to the Locations data structure for details.Locations
fashionModels	Yes	array[object]	Model information	1. You can obtain complete Locations information through the AI agent information query interface (apiReference#info) 2. Please refer to the Locations data structure for details.FashionModels
agentVersion	Yes	string	Agent version	/

Locations

location data structure

Parameter Name	Required	Type	Description	Notes
id	Yes	int	Location Id	/
image	Yes	string	Location Image	/
name	Yes	string	Location Name	/
categories	Yes	array[string]	Location Category	/
tags	No	array[object]	Location Tags	/

tag data structure

Parameter Name	Required	Type	Description	Note
id	Yes	int	Location Id（locationTagIds）	/
name	Yes	string	Location Name	/

locations

[
    {
        "id": 2113,
        "image": "https://ai-global-image.weshop.com/3e3fe7db-4cac-4c2d-8530-c4d9973b75ae_1024x1024.png_256x256.jpeg",
        "name": "Shopping Mall",
        "categories": ["Commercial Complex"],
        "tags": [
            {
                "id": 2034,
                "name": "Cinematic"
            },
            {
                "id": 2035,
                "name": "Film Grain"
            },
 
            ......
 
            {
                "id": 2036,
                "name": "Spot light"
            }
        ]
    },
    {
        "id": 2371,
        "image": "https://ai-global-image.weshop.com/9cba7891-dd9f-43ca-875d-1411cf9fd5ed_1024x1024.png_256x256.jpeg",
        "name": "Grocery Store",
        "categories": ["Store"],
        "tags": [
            {
                "id": 2012,
                "name": "Summer"
            },
            {
                "id": 2017,
                "name": "Fall"
            },
 
            ......
 
            {
                "id": 2004,
                "name": "Spring"
            }
        ]
    }
]

FashionModels

fashionModel data structure

Parameter Name	Required	Type	Description	Notes
id	Yes	int	Model Id	/
image	Yes	string	Model image	/
name	Yes	string	Model name	/
profile	No	object	Model informatoin, including country, style etc	/
type	Yes	string	Model Type: AI: Provided by WeShop AI Yours: Created by users, you may create your own AI model on the site	/
tags	No	array[object]	Model tags	/

profile data structure

Parameter Name	Required	Type	Description	Notes
nationality	No	int	Nationality of model	/
style	No	string	Style of model	/
priceDescription	No	string	Pricing of model	/

tag data structure

Parameter Name	Required	Type	Description	Notes
id	Yes	int	Model tag Id（fashionModelTagIds）	/
name	Yes	string	Model tags	/

fashionModels

[
    {
        "id": 20082,
        "image": "https://ai-global-image.weshop.com/92e45b1b-830d-4eaa-8ac2-340fc930230e_560x648.png_crop512.jpeg",
        "name": "Letitia",
        "profile": {
            "nationality": "Spain",
            "style": "Mature, Sexy",
            "priceDescription": "Additional cost of 5 points/image",
        },
        "type": "AI",
        "tags": [
            {
                "id": 2003,
                "name": "Smile"
            },
            {
                "id": 2014,
                "name": "Elegant"
            }
        ]
    },
    {
        "id": 20095,
        "image": "https://ai-global-image.weshop.com/a20ec42f-08b6-4b81-a4ea-36ce4a37fd2f_680x1024.png_crop512.jpeg",
        "name": "Eve",
        "profile": {
            "nationality": "Poland",
            "style": "Sweet, Cute",
            "priceDescription": "Additional cost of 5 points/image",
        },
        "type": "AI",
        "tags": [
            {
                "id": 2014,
                "name": "Elegant"
            },
            {
                "id": 2007,
                "name": "Laughing"
            },
            {
                "id": 2020,
                "name": "Fashionable"
            }
        ]
    }
]

AI Product

Create stunning product images with few clicks from anywhere

Basic Information

AgentName

aiproduct

AgentVersion

- v1.0

Note

When executing the task, please select at least one or more parameters from locationId , textDescription to execute.
- If you prefer to use the template, you can provide the locationId parameter. You can query the valid locationId values through the AI entity's info API.
- If you want to describe the image you want to generate with words/phrases/sentences, you can assign a value to the textDescription parameter.
- Mask settings:
  - If you only need to change the background without replacing the model's face, please set maskType to "autoSubjectSegment" when executing the task. The AI entity will use automatic subject segmentation technology to generate the mask image, and you do not need to provide customMask and customMaskUrl parameters at this time.
  - If you need to select the areas to keep unchanged on your own instead of using automatic segmentation, please prepare the mask image before the operation and set maskType to "custom" when executing the task. In this case, you need to provide either customMask or customMaskUrl parameter.

Tasks

Create Tasks

Request Parameters

initParams

Parameters Sample:

{
    "initParams": {
        "taskName": "Create Task Api Test",
        "originalImage": "https://ai-image.weshop.ai/xxxxxxxx.png"
    }
}

Parameter Name	Required	Type	Description	Notes
taskName	No	string	Task name	Do not exceed 20 characters
originalImage	Yes	string	Original Image	When using an external image link, you must ensure that the image link is publicly accessible on the internet.

Response Parameters

data

Parameter Name	Required	Type	Description	Notes
taskId	Yes	string	Task ID - the unique identifier	/

Execution

Request Parameters

params

Parameters Sample:

// Automatic Subject Segmentation
{
    "params": {
        "generatedContent": "referToOrigin",
        "maskType": "autoSubjectSegment", 
        "locationId": 2718,
        "textDescription": "A toy race car in front of a white wall.",
        "negTextDescription": "human faces",
    }
}
 
// Upload Custom Mask
{
    "params": {
        "generatedContent": "freeCreation",
        "maskType": "custom",
        "customMask": "<mask image base64 encoded>",
        "textDescription": "A toy race car in front of a white wall.", 
        "negTextDescription": "human faces",
    }
}

Parameter Name	Required	Type	Description	Notes
generatedContent	Yes	string	Select algorithm preference： "freeCreation"：Free creation "referToOrigin"：Refer to the original image	1. FreeCreation free creation will make the resulting image more free, without being constrained by the style of the uploaded original image. 2. referToOrigin reference original image will make the resulting image refer to the style of the uploaded original image during the generation process.
maskType	Yes	string	Select mask type: "autoApparelSegment": Automatic apparel segmentation "autoSubjectSegment": Automatic subject segmentation "custom": Manually upload mask information	1. autoApparelSegment automatic apparel segmentation is consistent with the automatic segmentation on the web page. 2. autoSubjectSegment automatic subject segmentation is for subject background segmentation, different from manual selection on the web page. 3. For manual upload, ensure the uploaded mask image has the same dimensions as the original image.
customMask	No	string	Upload the base64 encoded mask image	1. When the mask type is "custom", this parameter must be uploaded, otherwise no need to upload. 2. The area outside the mask region is transparent. 3. Use base64 encoding, without the prefix "data:image/png;base64". 4. Must have the same dimensions as the original image.
customMaskUrl	No	string	URL of the uploaded mask image	When using an external image link, you must ensure that the image link is publicly accessible on the internet. The image format requirements are the same as customMask.
locationId	No	int	Location	Optional
locationTagIds	No	array[int]	Location tags	Optional, please refer to Locations
textDescription	No	string	Text description of the content (optional)	Optional
negTextDescription	No	string	Description of content you don't want to appear (optional)	Optional
resultImageFormat	No	string	Select the format of the generated result image, options: "jpg" "png" (default is png)	Default is png
resizeToOriginalImage	No	bool	Whether to resize the generated result image to the original image dimensions	Default is false
batchCount	No	int	Set the number of images to generate in this execution, minimum is 1, maximum is 16	Default is 4, values out of range will use the default

Response Parameters

data

Parameter Name	Required	Type	Description	Notes
executionId	Yes	string	Execution ID - the unique identifier	/

Query Task

Request Parameters

json format, parameters sample

{
    "executionId":"<execution id>"
}

Parameter Name	Required	Type	Description	Notes
executionId	No	string	Execution ID	/
taskId	No	string	Task ID	If the taskId parameter is passed, at most the last 20 execution records will be returned.

Response Parameters

initParams

Current task's initial information

Parameter Name	Required	Type	Description	Notes
taskName	Yes	string	Task name	/
originalImage	Yes	string	Original image	/

executions

Execution information

Parameter Name	Required	Type	Description	Notes
executionId	Yes	string	Execution ID	/
status	Yes	string	Execution Status： Segmenting：In automatic segmentation Pending：Queuing Running ：In progress Success：Completed	/
params	Yes	object	Task execution related parameters	/
executionTime	Yes	string	Task execution timestamp	/
results	Yes	array[object]	Task execution result information	/

result

Execution result

Parameter Name	Required	Type	Description	Notes
status	Yes	string	Execution results Running Success Failed Pending	/
image	Yes	string	The generated final image	/
progress	No	string	Processing progress, e.g. 0.71 means 71% has been processed	/
error	No	string	Information if processing failed	When the status is "Failed", this field will definitely be returned.

result

Execution result

Parameter Name	Required	Type	Description	Notes
status	Yes	string	Execution results Running Success Failed Pending	/
image	Yes	string	The generated final image	/
progress	No	string	Processing progress, e.g. 0.71 means 71% has been processed	/
error	No	string	Information if processing failed	When the status is "Failed", this field will definitely be returned.

Information Query

Query information related to the AI agent. By passing in the agent name (agentName) and version number (agentVersion), you can obtain the corresponding configuration parameter information.

Request Parameters

Parameter Name	Required	Type	Description	Notes
agentName	Yes	string	agent name： "aiproduct"	/
agentVersion	Yes	string	agent version： "v1.0"	/

Response Parameters

data

Parameter Name	Required	Type	Description	Notes
locations	Yes	array[object]	Location information	1. You can obtain complete Locations information through the AI agent information query interface (apiReference#info) 2. Please refer to the Locations data structure for details.Locations
agentVersion	Yes	string	Agent version	/

Locations

location data structure

Parameter Name	Required	Type	Description	Notes
id	Yes	int	Location Id	/
image	Yes	string	Location Image	/
name	Yes	string	Location Name	/
categories	Yes	array[string]	Location Category	/
tags	No	array[object]	Location Tags	/

tag data structure

Parameter Name	Required	Type	Description	Note
id	Yes	int	Location Id（locationTagIds）	/
name	Yes	string	Location Name	/

locations

[
    {
        "id": 2718,
        "image": "https://ai-global-image.weshop.com/e2541c55-a6ef-4333-87a7-f2bdd5341a85_800x800.png_256x256.jpeg",
        "name": "White Wall Shadow",
        "categories": ["Light Effect"]
    },
    {
        "id": 2712,
        "image": "https://ai-global-image.weshop.com/59def4ad-5746-435a-991a-0ab83c190b2f_1024x1024.png_256x256.jpeg",
        "name": "Rooftop Terrace",
        "categories": ["Home Design"]
    },
 
    ......
 
    {
        "id": 2680,
        "image": "https://ai-global-image.weshop.com/9431f1dc-d40e-46ce-aebe-76c7a49e352b_1024x1024.png_256x256.jpeg", 
        "name": "Forest Window View",
        "categories": ["Natural Scenery"]
    }
]

API Reference Error Code