Goal
Read text from photo of tyre to determine size and type.
Progress
I copied and pasted the result into Chat-GPT.
I then used Chat-GPT to extract the info I need.
Here is the prompt I used:
https://chat.openai.com/share/117a3bff-596d-428b-ba9e-c25dae9dc5b6
I got the desired result in Chat which is great.
Help needed
How do I use the Integration / Open AI column to execute the same prompt?
Do you have a sample URL we can use to test? If you put the prompt in your share link to the system prompt in Glide, and adjust this part:
Below is some text extracted from a photo of a tyre:
<start>
A 185/55R16
83V TU
<end>
to something that references your tyre description input, would it work?
Also, why wouldn’t you just put the image as the input to save a column?
Do you have a sample URL we can use to test?
Do you mean a link to the Glide app? Or something else?
Here is the Open AI column:
Also, why wouldn’t you just put the image as the input to save a column?
Because I don’t know how to use this column properly yet 
Does the screenshot above show this now?
If I want to read text from an image (added to Image field) then do I just add a prompt (something like this?) to the prompt field?
What do I enter in the message field?
I would say “Prompt (system)” would serve mostly as persona definition for your “assistant”. I usually would just include the instructions in there as well, I don’t know if it makes a difference with the “message”.
So, I suggest adding your prompt in the chat link you share to system and see if it works without a message.
Thanks.
In Glide, Message field is required.
Hmm, so maybe get some of the instructions in there.
