OpenAI GPT-4 Vision: Custom Function
Build a MindStudio AI that leverages GPT-4 Vision to automatically describe what is in an image.
What is OpenAI GPT-4 Vision?
OpenAI GPT-4 Vision (GPT-4V) integrates image processing into GPT-4 by accepting images and answering questions about them.
What Can I Do With MindStudio and OpenAI GPT-4 Vision?
Build a MindStudio AI that leverages GPT-4 Vision to automatically describe what is in an image. With that text description, use another AI model to automatically generate ad copy for a marketing campaign.
For example, after receiving GPT-4 Vision’s response, use a Send Message Block to message an AI model to generate three different ad copy ideas based on the GPT-4 Vision’s text description of the image.
What is Required for This Custom Function?
OpenAI account.
OpenAI API Key. See Get OpenAI API Key.
GPT4-Vision Custom Function.
Get OpenAI API Key
Login to your OpenAI account.
Navigate to API Keys.
Select Create new API key.
Store the API key for the OpenAI API Key setting in MindStudio’s GPT-4 Vision Custom Function. Do not close the browser window until doing so.
Configuration
OpenAI API Key
Enter your OpenAI API key.
Image URL
Enter the image URL in which to generate marketing copy. Optionally, reference a Variable that stores:
the image URL (from a Short Text User Input) or
the uploaded image (from an Upload Image User Input).
Example: {{Image}}
GPT-4 Vision Instructions
Enter the instructions to send to GPT-4 Vision how to process the image. Example:
Optionally, reference a Variable that stores the instructions the app user enters into a User Input. Example:
Response Data Variable
Enter the Variable that stores GPT-4 Vision’s response. Example:
Success Variable
Enter the Variable in which to store that GPT-4 Vision processed the image successfully. Note to not use double curly braces when entering the Variable name into the Success Variable setting. Example:
Last updated