Hi ​@vani.naveen,

Welcome to the Community.

To troubleshoot image generation issues in ChatGPT, first verify that you’re using the correct API version that supports this functionality, such as version 2.36.1. The ChatGPT 3.5 Turbo Instruct model mainly handles text interactions and may not support direct image generation. If using an image-related API, check for any limitations or quotas in OpenAI’s documentation. Reviewing error details or logs can provide further insight into resolving the issue.

I hope this helps. Please let me know if there's anything else I can assist you with. We're here to help.

Okay. Thank you for the info. Even for image generation we need to check the ChatGPT model. Because earlier there was no error encountered while generating. 

Hi ​@vani.naveen,

If you're encountering new errors with image generation using a ChatGPT model, consider recent changes to your setup, such as model updates or input modifications. Check if you've hit usage limits or quotas, and review error messages for clues. Testing with different prompts can help isolate the issue, and consulting the latest documentation may reveal updates or required configuration changes.

If you have any further questions or need additional assistance, please feel free to ask. We're here to help you resolve any challenges you encounter.

You mean the model change and image prompt. Yes i have changed image prompt but model not changed. i check still Quota or limit is there in the account. Still same error persists. Let me check by changing the model.

Hi ​@vani.naveen,

I believe the issue here is due to the prompt exceeds the 4000 character limitation. Could you please try writing a prompt that’s 4000 characters or less?

I'll be keeping an eye out for your response!

i tried with less characters still same error so yesterday i tried by changing the model for chatgpt and tried one image it generated

Hey there, ​@vani.naveen 👋

Hope you don’t mind me jumping in here. It sounds like you were able to get it working by both reducing the length of the prompt as Ken suggested and changing the model that was selected in the Generate An Image action, is that correct? If so, can you confirm which model you switched to to get it working?

The help text for the Prompt field states that the dall-e-2 model has a character limit of 1000 and the dall-e-3 model has a character limit of 4000:

And you mentioned you were using ChatGPT 3.5 turbo instruct initially but I’m not seeing that listed as one of the available models when I test in my own Zap:

Looking at your first screenshot it seems you may have been using an older version of the ChatGPT app as there’s a version number appearing next to the app name:

So perhaps that’s why I’m seeing different models available as I’m using the latest version of the app.

Are you still using that older version of the app or has it been automatically upgraded to the latest version? If it was upgraded, are you now seeing the same model options as the ones I’m seeing or do you still have access to ChatGPT 3.5 turbo instruct? Just wondering if the app version could have also had a part to play in the cause of the error? 🤔

Looking forward to hearing from you! 

Yes when i connected with API KEY the ChatGPT model says 2.36.1 earlier and send prompt model i have used ChatGPT 3.5 turbo instruct and to image generation the model we got to select dall e- 2, dall e 3 and gpt image 1 but i have used dall e 3 for all the images we generated till now. 

and recently i have changed the model i have used is chatgpt 3.5 turbo instruct -0914 still same error persists.

As we are creating final zap to take full automation we need featured image and section images for the complete blog we were working that time the error halted like this.

Hi ​@vani.naveen,

If you're facing issues integrating ChatGPT and DALL-E models via API in your Zapier workflow. First, verify that your API key is correct and has the necessary permissions. Ensure model compatibility with your API version, checking documentation for updates. If errors occur, providing specific details will help diagnose the problem. Confirm that your Zap configuration is set up correctly with proper data formatting. Be mindful of API rate limits, and consider testing each model separately to pinpoint where the issue lies.

Feel free to reach out if you have any questions or need further assistance.

Hey ​@vani.naveen, just checking in—were you able to get this sorted? 

Keep us posted on how it’s going, want to make sure you’re all set! 🙂