Best 'text-to-speech' api to plug into a chatgpt bot?


I was wondering if anyone had any good ‘text-to-speech’ options for use with browser based chatgpt bots.

  • any examples out there


I am also looking for this type of application. I’ve been looking for it, trying to find it, but given the number of hours of text I have to transcribe, I end up either with offers that are not capable of doing it, around 100,000 characters for 20€ for example, or with offers at exorbitant prices.
To give an example, this is an American tutorial that I want to transcribe into French. There are about 40 videos of 50/60 minutes each.
A 50-minute video has about 40,000 characters.
So with a quote of 22€ I will make about 2.5 videos.
That comes to 350€. It’s not the end of the world, but let’s say I could get away with 50/100€, that would be more acceptable. Do you have any ideas? Even if I have to use paid APIs to stay within that price range?

I have been using voice models for my voice conversation feature for my project, linked in my bio.

It depends if you care about price or quality or latency.

I have been using Google Voice and that one works quite well and they have options for cheap but not great sounding or expensive but better sounding. All of those are quick enough to be almost real-time.

Elevenlabs sounds the best imo but much more expensive.

There are other players coming around, Bark is open source but the quality is not that great IMO. Charactr seemed interesting but I haven’t tried it yet.

There are also browser native tts libraries out there but they don’t really sound great.