I am trying to automate pictures of handwritten forms into a spreadsheet, the only issue is that gpt-4-vision-preview has been deprecated. The docs say to use gpt-4o instead.
>model_not_found] The model `gpt-4-vision-preview` has been deprecated, learn more here: https://platform.openai.com/docs/deprecations
Any way to fix this so I can use the default analyse image content with vision function in Zapier?
+14
Hi
In your Zap step, you’ll need to update to use an available ChatGPT model.
+14
Did you try changing the model as suggest to:
gpt-4oOr try using this Zap action: ChatGPT - Analyze Image Content with Vision
+14
I just tested and it worked. (screenshot)
The Image field description states: GPT-4 access is required.
Do you have GPT-4 access via a ChatGPT+ subscription?
Help articles for using ChatGPT in Zaps: https://zapier.com/apps/chatgpt/help
+14
Are you using the most updated version of the ChatGPT Zap app integration? (top option)
+14
Are you passing in the URL of the image(s) as indicated in the Image field description?
+14
Can you post a screenshots with how your Zap step is configured in EDIT mode for the ‘Action’ tab section?
+14
Check if the link is for a supported image format.
Check if the link is publicly accessible.
+14
You can try opening a ticket with Zapier Support for possible help: https://zapier.com/app/get-help
+11
Hi
I did some checking and according to the OpenAI’s documentation at the link the error references, the gpt-4-vision-preview model isn’t being deprecated until December 6th, 2024, but it also mentions that only existing users of these models will be able to continue using them until that shutdown date:
So if you weren’t already using that gpt-4-vision-preview model before the announcement they made on June 6th then I expect that’s why you’re running into this error.
The Analyse Image Content with Vision action will be switched over to use the gpt-4o model ahead of that shutdown date but I can’t provide an estimate on when that will be exactly. Once it is that should help to prevent those errors you’re currently seeing and allow you to use that action.
In the meantime, perhaps you could try using the Conversation With Assistant (ChatGPT) action, or Conversation (ChatGPT) action, to analyse the images instead? Both have the option to allow you to select the gpt-4o model and have file/image fields that you could use to pass it image files that you want it to analyse.
Can you give that other action a try and let me know if that works as hoped?
+11
Hey there
I just wanted to follow up here with a quick update to share that the team have updated the Analyse Image Content with Vision action to use the gpt-4o model so you should no longer be seeing those “[model_not_found] The model `gpt-4-vision-preview` has been deprecated...” errors. 😁
If you run into any further issues or have any questions please do let us know. In the meantime, happy Zapping! ⚡
Enter your E-mail address. We'll send you an e-mail with instructions to reset your password.
