When will GPT-4V and DALL-E 3 be available?

When will GPT-4V and DALL-E 3 be available?

The papers are interesting but access is better!

We need access for a project choice before making a decision.


Besides the announcement that developer access is in the making we have no timeline as to when and how this will happen.
You could likely wait for the developer conference on 6th November, expecting that they will clarify the rollout timeline.
Also, nobody really knows if this may be another waitlist situation.
Let’s just hope for the best!

Thank you for the information. We’ll just have to wait then which is most unfortunate because the papers are out. Early access should be available for developers/authors preparing publications and advising entities.

It is being released in batches. I received access 2 weeks ago. This Monday I noticed a massive distribution in Brazil, as different friends received it on the same day.

Just to confirm: Are you referring to API access?

1 Like

Thank you for the information.

I’ll watch for access which will hopefully arrive soon.

Thank you.

I’m referring to both online and API access.

Is online access available somewhere, please?

How do we know when API access is available?

My reply was meant to be for @natanael.wf
I am very confident that API access is not yet available but maybe I am wrong?


You are correct, there is currently no DALL-E-3 API, I would expect it by the end of the year, perhaps sooner.

1 Like

I am hoping for DALLE-3 API to be released at DevDay.

You are correct, I do not know of anyone with API access and the OpenAI api docs don’t reference any way to use images as inputs to chat completions.

1 Like

I was referring to UI access, not the API.

Have you already activated the Beta Features in your ChatGPT settings?

Yes. It seems that the DALL-E 3 API is not available yet.

The model is only available in ChatGPT.

Yes. I activated the beta features in ChatGPT Plus.

DALL-E 3 is available.

GPT-4V isn’t.

It’s odd, everyone I know got access to both features simultaneously. Sorry to keep pressing, but have you noticed that when you select the ‘Default’ model, GPT-V is, in practice, just an image icon to the left of the ‘Send a message’ field?

This means you have GPT-4V


Don’t be sorry! Your message is very helpful. I was just expecting something more spectacular.
But I was wrong.
The fundamental spectacular innovation is that GPT - 4V is both an LLM and an advanced CV with integrated capabilities and cross-attention.
It’s rather exciting in terms of its architecture. :blush:

Thanks for your helpful guidance!

1 Like

Thank you for your response.

Yes. This works fine for me.

I was just looking for something spectacular at the UI level.

However, the real extraordinary fact is the architecture of this true foundation model that is a great LLM and CV all in one.

Thanks for your help.

1 Like

Can you clarify as I’m missing something?
ChatGPT-4 has been available for a long time and I added access via API in our software many months ago. I’m also running multiple other local LLMs - LLAMA2 and Falcon are the two we currently use in our GP practice.
And Microsoft via Bing is now running ChatGPT-4 as it’s response engine as well as offering access to DALL-E.
So I don’t understand what the question is actually asking.

However I do have a problem with accessing them from an NHS login - which seems to be blocking for some reason (IT have denied it’s them but since they have blocked the OneDrive app I don’t currently believe them). Is that what you mean?