GPT 4 Vision - Multiple Images to Text

Hey! I have a problem to make the API with GPT-4 Vision work.

I am looking to add multiple pictures of properties into an image uploader that then will generate a description of these images in another group.

My API Connector is initializing correctly. Here is the JSON:

{
“model”: “gpt-4o”,
“messages”: [
{
“role”: “user”,
“content”: [
{
“type”: “text”,
“text”: “Provide a detailed description of the property based on the provided images and the following additional information: number of bedrooms, number of bathrooms, number of parking spaces, and property address. The description should be clear, attractive, and SEO-optimized, highlighting each area of the property and nearby amenities, without including fictional or unverifiable information from the images.”
},
{
“type”: “image_url”,
“image_url”: {
“url”: “<uploaded_image_1_url>”
}
},
{
“type”: “image_url”,
“image_url”: {
“url”: “<uploaded_image_2_url>”
}
},
{
“type”: “image_url”,
“image_url”: {
“url”: “<uploaded_image_3_url>”
}
}
]
}
],
“functions”: [
{
“name”: “analyze_property_and_generate_description”,
“description”: “Analyzes property images and generates a detailed description based on the social areas, private areas, service areas, leisure spaces, and location, always considering the provided inputs.”,
“parameters”: {
“type”: “object”,
“properties”: {
“general_description”: {
“type”: “string”,
“description”: “General description of the property, including size, type of property, and initial highlights such as location and view.”
},
“living_area”: {
“type”: “string”,
“description”: “Description of the social area, such as living room, dining room, and balcony, mentioning finishes and views as shown in the images.”
},
“bedrooms”: {
“type”: “string”,
“description”: “Description of the bedrooms, including the number of rooms and their features, such as built-in closets, lighting, etc., based on the images.”
},
“bathrooms”: {
“type”: “string”,
“description”: “Description of the bathrooms, including finishes and features such as bathtubs, as shown in the provided images.”
},
“kitchen_and_service_area”: {
“type”: “string”,
“description”: “Description of the kitchen and service area, including cabinets, appliances, and dining space, based on the images.”
},
“parking”: {
“type”: “string”,
“description”: “Description of the number of parking spaces and other relevant details observable in the images.”
},
“building_amenities”: {
“type”: “string”,
“description”: “Description of the leisure areas and building conveniences, such as the pool, gym, and other amenities observed in the images.”
},
“neighborhood”: {
“type”: “string”,
“description”: “Description of the location and nearby amenities, based on the provided address, mentioning shopping malls, supermarkets, and other points of interest if precise information is available.”
}
},
“required”: [
“general_description”,
“living_area”,
“bedrooms”,
“bathrooms”,
“kitchen_and_service_area”,
“parking”,
“building_amenities”,
“neighborhood”
]
}
}
]
}

My Workflow is as below:




And the message error I have is:


→ I tried to use a Base64 encoder (from ZeroQode), but could make it work.
→ how do I send multiple images to the GPT-4 Vision API through the Image Uploader that is able to receive all the images I sent him? and in which format?

Any help is welcome! Thanks in advance to those that can help me with that.

Ben