Use the Free version of Prime Voice AI?
Opened this issue · 1 comments
Just wondering if this would add some benefits over the current Coqui-ai's vits model?
The beta can be found here.
Hi @bornagainpenguin!
Thanks for this. On a first glance, it looks really cool. Let me try and test it out locally. They are saying that 10000 characters should give us about 10 minutes of voice over.
My back-of-the-envelope calculation tells me that 10,000 characters should give us about 1500 words or 3-4 pages. One normal sized schema should take around 3/4th page (assuming that we synthesize news as well). So, we might be able to get 4-5 highly quality radio broadcasts out of this. Maybe we can then default to Coqui-ai.
On another note:
I was wondering if you know of any good open-source alternatives to Coqui-ai? I've heard that Mimic 3 from Mycroft and tortoises-tts are good, but Mimic 3's voice seems a bit too robotic and tortoises-tts seems awfully slow on CPUs. I am trying to understand how much we could push this project towards OSS-based TTS before we have to integrate proprietary solutions.