Your Voice. Your Phone. Your Privacy.
ToBe SAID Lite is a high-performance offline AI voice generator and system-wide Text-to-Speech (TTS) engine built for total privacy. Unlike cloud-based voice apps, everything runs 100% locally on your device — no servers, no subscriptions.
Why choose ToBe SAID Lite?
🔒 Essential Privacy: Process all speech and voice synthesis offline. No accounts or internet required.
⚙️ System-Wide Integration: Power TalkBack, screen readers, and eBook apps with high-quality preset voices.
🗣 Single Custom Voice: Create one personal AI voice from your own recordings. Adding a new voice replaces the old one.
🚀 Low Latency: Optimized C++ and NDK integration ensures fast speech generation without the cloud.
Who is it for?
Accessibility Users: Upgrade TalkBack with natural AI voices that respond instantly.
Mobile Readers: Power your favorite eBook and PDF apps with premium offline narration.
Privacy Advocates: Ensure your biometric voice data remains strictly on your local hardware.
Important Note: This Lite release supports English only. High-priority updates for multilingual support are in development.
Note: This version is limited to 1 custom voice slot. For unlimited slots and custom voice TTS, see the Pro version.
View on Google Play StoreFree or paid:
- Free with in-app purchase
Accessibility and User Comments:
It’s no secret that over the years, the choice of TTS engines on Android has continued to shrink, leaving only a few options on the Play Store.
For those who prefer sideloading APKs, there are some TTS engines available on GitHub. However, finding these APKs can be difficult for those who are not familiar with the interface. As a result, when a developer makes an app available on the Play Store, such a developer automatically receives glowing praise from me!
This is the case for “Lookbe,” whose app is the “ToBe Said” TTS engine.
Upon installing and opening the app, the typical “Notification” access dialog appears. This can be allowed or denied.
The main window of the app includes the following items and controls:
- “Read clipboard” – inserts copied text into the text field for reading.
- “System text to speech settings” – opens the device’s TTS settings where “ToBe Said” can be selected.
- “Recite Hamlet” – pastes a predefined text into the edit field.
- Edit field – displays the pasted text or allows entering custom text.
At the bottom, there are three buttons from left to right:
- “Open settings” – opens the app settings.
- “Select voice” – opens the list of available voices to choose from.
- “Speak text” – speaks the text entered in the edit field above.
Settings
The settings screen includes the following options:
- “Buy Pro Version” – allows generating multiple voices and using them as system voices (coming soon).
- “Theme” – opens the theme selection screen to choose a preferred theme.
- “Share Mode” – offers a choice between compatibility and quality.
- “Voice Quality” – allows selecting “Low,” “Medium” (default), or “High Quality.”
- “System text to speech settings” – opens the device’s system TTS settings.
Other items such as reporting bugs, viewing terms, licenses, and the app version, are also available in the settings.
Choosing a VOICE
To choose a voice variant other than the default, tap the “Select Voice” button, which also displays the name of the currently selected voice. Tapping it opens a window with the available voices.
Below each voice on the left is “Preview,” which plays the sample voice when tapped.
To choose a voice, tap its name and it becomes the default.
At the bottom of the voices screen, there are two buttons:
- “Upload” – allows selecting a prerecorded voice to train a custom voice that can be added to the available voices.
- “Record” – allows recording a voice sample used for generating a voice variant.
Accessibility
When it comes to accessibility, no major issues were found in the app itself. The interface is simple and not complicated.
However, the TTS engine is not yet ready to be used as a daily screen reader voice. There is noticeable latency, and sometimes it may sound as if there is buffering during speech because audio is not generated quickly enough, which affects the overall quality.
The app is receiving updates with attempts to improve overall stability and responsiveness, though.
The ability to generate voices and use them as system voices is interesting and worth monitoring to see how well it performs when fully functional.
Audio Demonstration
You can listen to my audio demonstration of ToBe SAID here.
Last Tested App version and Android version:
App version: 1.0.5, Android 16

But does this TTS read emojis? I’m disappointed by its inability to respond to talkback’s speech rate.
Hello, I’m the developer of the app. I’m still fixing talk back issue, the fastest I can get is next month for the update. And thanks for reporting the emoji issue
There has been a couple or so updates that now makes it possible to change the speech-rate. The demo was recorded before these updates.
Eu acho bem melhor o jieshuo tinha que reativar denovo