What is your favorite chatGPT Voice?

Hey champ, and welcome to the community,

these specific voices aren’t available for commercial use, but as mentioned, there’s quite a few other’s available though the API that you can use :laughing:

2 Likes

I will repeat this louder in case it was overlooked:
If I had only one request, it would be for the volume to be louder. In the Android app, even with the volume at max, I can barely hear Whisper speak.

1 Like

I love the sound and accent of Ember. After that of breeze, i like it. I love to hear ember so i voted for it. Do u know which accent ember speaks in?

2 Likes

Unfortunately not, I’m not from the US, but maybe some of our American friends can chime in here? :smiley:

Ember isn’t really an accent, sounding like normal broadcast English like you’d hear from New York to California, but the pronunciation is overemphasized, like it is someone doing a radio audition.

Then you have Cove, who sounds like he was just woken up.

1 Like

Any plans to release these Chat-only voices for API use in the near future? I really, really like Ember, and it’s much better than all the current API voices atm imo. The emotional range retains my attention longer than the others.

1 Like

Are these voices not accessible via API? (talking about Sky, Juniper and the likes)

I checked the docs and the basic tts model voice options are the only ones available.

They are not; the API has different voices available.

We did ask OpenAI about it, and the voices that are available through ChatGPT are created in cooperation with real voice actors, while the ones available through the API are synthesized from many different people.

The contract OpenAI has with these voice actors makes it so their voices are only available through ChatGPT.

2 Likes

Hi, I have a question regarding the voice actors. My brother has told me yesterday that the Cove voice sounds remarkably like mine. I work as an actor and do voicework myself. I thought it was incredibly similar as well, along with his wife and several of my friends, some of whom work as directors or sound engineers.

Of course I’m not here to point fingers or make accusations, but it has made me quite uncomfortable, seeing videos about the software, TikToks and the like and I can’t unhear it. Do we have any verifiable information on the Voice Actors used to train these AI models? Or perhaps any advice on how to go about asking and the like? I even have clips of me and the AI reading out the same text and it’s honestly a bit uncanny, to me and several others. Can I find help on the issue, please?

Hey @YASON and welcome to the developer community forum!

As previously mentioned, OpenAI has a contract with the voice actors who’s voice’s where used to create the voices you hear while using ChatGPT. Therefore, your voice was not used in any way.

2 Likes

how do we use the new voices that were in this weeks demo? They sounded much more conversational that the ones popping up on the app right now. Are they just not live yet?

Yep, these voices are not currently available.

My understanding is that those are not “voices” in the same sense as the other TTS voices we have.

The model is natively multimodal, so a different voice would be an entirely different model.

The voices you heard in the demonstration are the intrinsic voices of the GPT-4o models.

3 Likes

I really miss the ability to use Juniper via the API.