I’ve created a Form Pickaxe and enabled the ability for a user to upload a file.
In my example it’s a headshot photo.
My prompt uses two user inputs
- A text field that allows for a description of an image
- A radio button with image styles
How do I get the prompt to take the uploaded image and use it as part of the output?
For example…
Prompt: Take the uploaded image of a headshot and {User Image Description} and {Image Style} and produce an image.
Form
User Image Description: Create a classic Cluedo character, with the solid colour plastic body with a head on top. Name of the character is “My Name”
Image Style: Classic Cluedo Game Card
In the prompt I can reference the user inputs for the “image style” (radio button) and “user’s image description” (long text field) but I can’t see a way to target the uploaded input?
Any ideas?
1 Like
Uploaded files (like images) typically aren’t passed directly into the prompt as usable input. Instead, they’re stored and made accessible via a temporary file URL — but that URL doesn’t get automatically injected into the prompt like other form fields.
Just to better understand your setup:
- Are you using a model that supports image input (like GPT-4 with vision)?
- Are you hoping the model will analyze the uploaded image, or simply use it as a reference to generate a new one (e.g., with DALL·E or a similar tool)?
Happy to help once I know a bit more!
I was going for the latest ChatGPT Image.1 model
I was looking for it to take the image and the prompt and create a new image that included that as the reference. I.e. take the person’s headshot and add them into this other type of image. Making sure the head is placed in a specific position.