I’ve been thinking about upgrading from Premium to Premium Plus. The main thing pulling me in is the AI-generated voices. I’ve heard Steve talk about how good they are in his videos (at least I think that’s what he was referring to), and I’m hoping they’re a big improvement.
The regular audio generation on Premium can sound pretty harsh on the ears sometimes, especially during longer listening sessions. So I’m mainly wondering if the ElevenLabs AI voices actually make a noticeable difference in quality and comfort.
If anyone here has Premium Plus and regularly uses the AI voices, I’d really appreciate your honest take — is it worth the extra cost for you? That’s pretty much the only new feature I’m interested in right now.
It depends on the language you’re learning, but with the latest added voices, for me, it’s only French that still has a premium voice that is anywhere near the quality of the plus voices.
The added audio transcription time is also a big bonus with the improved uptadates of it.
Thanks Zaya, I was about to post the exact same question. I generated audio for a book chapter in spanish today, and I can understand it, but the lack of any natural pauses and the robotic nature of the voice are distracting.
I wonder if the 11 labs versions are more comfortable to listen to / less distracting. Has anyone used the Premium plus for Spanish audio generation?
I suppose I may have to try for a month and see. If I do, I’ll report back to you @ZayaFTW
It’s not really about the money. It’s about whether or not it’s a good quality audio generation for each lesson.
I had a look at the audio quality for Russian here[1]and German here[2] and both from Eleven Labs sounded very good. Some voices sound better than others.
Yes the quality is terrible on the audio generation and the sharp sounds are very harsh on the ears if played at a loud volume.
I found their website and everything about the website screams quality. I tested Russian and German, and both sound great or at least much more natural sounding than the current Premium tier audio generation.
Yes. This is the generate audio for lesson feature which, is useful for when you have text without audio to match. And it will synchronize the reading of the audio with the text.
For mandarin in its current inplementation: No. Absolutely not.
In general: Depends on how much you value having more natural sounding tts. Honestly think its much more worthwhile to use native audio sources for listening practice. Even generated audio podcasts that aren’t reading a set script, like notebookLLM, sound a lot better than the current tts implementation imo.
