Image Query

This function allows you to query an image or a set of images with a prompt. Save the requestId from the response and use the status endpoint to check the status of the request and retrieve the output. The imageUrls is an array of one or more public URLs of the images to query. You can use the assets/create endpoint to upload your images. To get the public URL of the uploaded image, simply remove the parameters. eg. everything after and including the ? in the returned assetUrl from the assets/create response. Make sure you have uploaded the image to our servers before querying.

API Reference:

POST /llm/image_query

Function Signature

client = Masterpiecex()
client.llms.image_query(**kwargs) -> GenerateResponseObject

client = Masterpiecex();
client.llms.imageQuery(body, options?): GenerateResponseObject

Parameters

Python	Node	Description
user_prompt: str	userPrompt: string	The user prompt to use for the LLM call to analyze the images
image_urls: List[str]	imageUrls: Array<string>	The list of public urls of the images to query. Should be an array of strings.

Returns

GenerateResponseObject

Example

import os
from mpx_genai_sdk import Masterpiecex

lient = Masterpiecex(
    bearer_token=os.environ.get("MPX_SDK_BEARER_TOKEN"),  # This is the default and can be omitted
)
generate_response_object = client.llms.image_query(
    user_prompt="What can you tell me about these images",
    image_urls=["https://,,,image1.png", "https://...image2.jpg]
)
print(generate_response_object.request_id)

import Masterpiecex from 'mpx-genai-sdk';

const client = new Masterpiecex({
  bearerToken: process.env['MPX_SDK_BEARER_TOKEN'], // This is the default and can be omitted
});

const generateResponseObject = await client.llms.imageQuery({
    userPrompt="What can you tell me about these images",
    imageUrls=["https://,,,image1.png", "https://...image2.jpg]
});

console.log(generateResponseObject.requestId);

Example Response

print(response.outputs.output) # The response from the llm call
print(response.balance)  # remaining credits available associated with the account
print(response.request_id) # used to check the the status. Eg., client.status.retrieve(request_id)
print(response.status)  # current status of the request - typically pending on initial submission

console.log(response.outputs.output); // The response from the llm call
console.log(response.balance); // remaining credits available associated with the account
console.log(response.requestId); // used to check the the status. Eg., client.status.retrieve(requestId)
console.log(response.status); // current status of the request - typically pending on initial submission

{
    "requestId": "xxxxx",
    "status": "complete",
    "processingTime_s": 5.777,
    "outputs": {
        "output"="The first image is a 3D rendering of a cartoon-style dog with exaggerated features like large eyes and a big nose, giving it a playful and cute appearance.\n\nThe second image is a 3D rendering of a humanoid robot. It has a sleek, white design with jointed limbs, making it look modern and futuristic. The robot's design suggests it could be used for various interactive or educational purposes."
    }
}