Why is gpt-4o model not able to generate images on Azure AI Studio?

Tanaka, Marcela 25

Hello!
The newest model GPT-4o was made available for deployments on Azure AI Studio, however, it is missing important features that are present on the OpenAI's version, such as image interpretation and generation.

Can someone help me to understand why?

Thanks in advance.

navba-MSFT 18,725 Reputation points Microsoft Employee

2024-05-23T03:13:59.0733333+00:00

@Tanaka, Marcela Welcome to Microsoft Q&A Forum, Thank you for posting your query here!

.

Image interpretation:
GPT-4 Turbo with Vision is a large multimodal model (LMM) developed by OpenAI that can analyze images and provide textual responses to questions about them. It incorporates both natural language processing and visual understanding. More info **here about the quickstart.
**
For example:

.

Image generation:

Azure OpenAI Service offers a model called DALL-E. This model can generate original images from natural language descriptions. The image generation API creates an image from a text prompt. It does not edit existing images or create variations. More info here.

.

Hope this helps. If you have any follow-up questions, please let me know. I would be happy to help.

**

Please do not forget to "Accept the answer” and “up-vote” wherever the information provided helps you, this can be beneficial to other community members.
Tanaka, Marcela 25 Reputation points

2024-05-23T12:42:32.6066667+00:00

Hi! Thanks for the very thorough answer!

However, I was referring to the new model GPT-4o. This model was built to be end-to-end, accepting several kinds of input (text, audio, video, files) and also generating several kinds of output.

When I tried to use it on Azure AI Studio, it was able to interpret the image, however, it was not able to generate any images. The example is attached below.

Is it correct to assume that the GPT-4o available in Azure is not the same as the one on OpenAI regarding multimodal capabilities?
navba-MSFT 18,725 Reputation points Microsoft Employee

2024-05-23T15:14:31.3566667+00:00

@Tanaka, Marcela Thanks for clarifying your ask. I am checking this and I will get back to you with an update.
navba-MSFT 18,725 Reputation points Microsoft Employee

2024-05-27T06:18:53.8366667+00:00

@Tanaka, Marcela Just following up to check if the below answer helped. If that answers your query, do click "Accept the answer” for the same, which might be beneficial to other community members reading this thread. And, if you have any further query do let me know. I would be happy to help.
Nehorayoff, Andre 5 Reputation points

2024-06-12T13:03:56.7333333+00:00

Bumping this request up - @navba-MSFT have you had a chance to confirm why image and audio generation functionality for GPT-4o (Omni) is not yet in Azure OpenAI Studio? Will this be available via the API?

Accepted answer

navba-MSFT 18,725 Reputation points Microsoft Employee

2024-05-23T17:18:00.06+00:00

@Tanaka, Marcela Yes, Your assumption is correct.

.

Azure OpenAI GPT-4o model will interpret images but not generate one at this time. If you want to generate the image you need to call Dall-E as I mentioned in my above comment.

.

Hope this helps. If you have any follow-up questions, please let me know. I would be happy to help.

**

Please do not forget to "Accept the answer” and “up-vote” wherever the information provided helps you, this can be beneficial to other community members.
Please sign in to rate this answer.
navba-MSFT 18,725 Reputation points Microsoft Employee

2024-05-27T06:25:32.01+00:00

@Tanaka, Marcela Just following up to check if the above answer helped. If that answers your query, do click "Accept the answer” for the same, which might be beneficial to other community members reading this thread. And, if you have any further query do let me know. I would be happy to help.
Sign in to comment

Share via

Why is gpt-4o model not able to generate images on Azure AI Studio?

0 additional answers