Hello AppleVis community,
I'm Meriç Karaca, the developer of ZeroTTS. I'd like to invite VoiceOver users to help test the app through TestFlight. My aim is to give blind and low-vision users more choice in the voices they listen to every day, while improving responsiveness, clarity and accessibility with feedback from real use.
ZeroTTS provides a selection of male and female voices, with support for 31 languages. Speech is generated on the device, and the voice models are included with the app. You can listen to samples inside ZeroTTS before choosing which voices to enable for VoiceOver. There are also adjustable speech-quality and pause-length settings. The app requires iOS or iPadOS 18 or later.
To get started:
1. Install TestFlight and join the ZeroTTS beta using the link below.
2. Open ZeroTTS, go to Voices, and enable the voices you would like to try.
3. Open Settings > Accessibility > VoiceOver > Speech, choose your language, and select a ZeroTTS voice. Depending on your iOS version and setup, you may add it through Add Rotor Voice.
4. Try it during everyday VoiceOver navigation as well as with the app's built-in samples.
This is an actively developed beta. Response delays, pronunciation issues and incomplete word endings are still being investigated; results can vary by voice, language and device. A voice sounding good in a sample does not necessarily mean it will work equally well when navigating quickly with VoiceOver. That is an important part of what I would like to test.
Feedback would be especially helpful on:
- How quickly speech begins when you move to a new item, compared with returning to an item that has already been spoken.
- Whether words and sentence endings are read completely, including names and app names.
- Individual letters while typing, numbers, dates and times in your language.
- Pauses, speech at faster rates, audio quality and any unexpected sounds.
- Whether enabled voices appear in VoiceOver's list, and how the app performs on older iPhones and iPads.
Please reply here or send feedback through TestFlight. Include your device model, iOS/iPadOS version, ZeroTTS build, selected voice and language, speech rate, and quality/pause settings when relevant. A short non-private example of the text and what you heard would help make pronunciation reports reproducible.
Thank you for helping shape ZeroTTS around the needs of people who use VoiceOver every day. Testers in all supported languages are welcome!
Disclosure: This post was drafted with AI assistance using information from my ZeroTTS development work.
Join the ZeroTTS beta on TestFlight:
https://testflight.apple.com/join/n2ZfJ2Sa
Comments
Feedback
Hello,
Thank you for your synthesizer!
Here is my feedback about first usage on iPhone 13 running iOS 26.6.1:
I tried using voice Tuana and found out that synthesizer has big pauses before starting to read the sentence or an icon on a home screen.
It doesn't read numbers correctly as a lot of AI voices.
It has pauses when switching between languages in the message.
Maybe it is so slow because of my iPhone model, I am not sure.
Do I understand correctly that there is no way to choose language and all voices speak all 31 languages as multilingual voices?
Thank you!
HarmonicaPlayers opinion
i chose to enable all voices in the app and i think the samples sound good :) :) i use the app called text to speech that allows you to type text and export it to mp3. i'll be using 0tts alina for one of the voices cause that one sounds nice :) and a few others do i'll be using eventually to :) also i have an iPhone 17 pro max running latest ios stable version
You are on a Role.
do you have a social media page?
You're on a role with ∫ apps.
Thanks.
Do you think you'll build a camra app?
I need a good accessible one.
I need it to have proffessional knowledge about photography.
1. Tells me what's in the view finder and asks what I want to do.
2. understands what I'm doing it for so it inteligently tells me what to do, such as,
Move the camra right,
Tilt the phone upward.
3. inteligently sets the right settings for me because I obviously am blind but also I'm not a photographer so I don't know them well enough to do it myself.
4. once it tells me the view is perfect for my needs, it will take the photo for me.
The problems this would solve.
I used to have to screen shot the view,
then share to be my AI.
I'd then have to use 2 hands to type in the BSI what I need the photo for so what should I do.
where does the phone need movement and what should the settings be on.
are there things in the photo I don't want there?
By the time I did that my hands had adjusted the phone and it is difficult to get it back that way.
Even with IOS 27,
I would get a quicker description yes but to tell it what I want after and ask it if the views good means the problem before is still not gone.
I need to tell it and give it that context before I take the photo.
I then need to get the view write and adjusting my fingers and swiping to the take picture button makes me loose that view again.
is that what focus lock is for?
Anyway It would still just be easier if it just took the photo once it's how I want it but what's proffessional.
If you could make that it would be so useful.