New features and improvements made in Braina version 2.5 are as follows:
Local Text to Speech Voice Models
Until now, only the default voices that came with the Windows operating system have been available in Braina for offline use. This new version of Braina supports numerous high-quality offline voices for various languages.
A list of available offline voices can be found here: Offline TTS Voice Library. You can download offline voices by going to Settings -> Voice & Language tab, clicking on the “Download Offline Voice” button and then entering the model tag for the voice model to download. Once the voice model is downloaded, you can select it from the “Select Offline Voice” dropdown and click on Ok button to apply changes.
Note: To use offline voices even when you are online, please select “Disabled – Use Offline Voice” in online voice selection. To use a non-English voice, enable multi-language mode and select your preferred output language in Braina to view the voices that you have downloaded for that language in the Select Offline Voice dropdown menu.
You can delete a downloaded offline voice model, by clicking on the “Delete Offline Voice” button, selecting the voice to be deleted and then clicking on the “Delete” button.
Offline voices are available in multiple qualities: Medium, HD, Lite, and X-Lite. HD voices have the best quality but requires pro version.
Offline voices, especially the HD models, sound more natural and human-like. However, the online “advanced AI voices” are still the best voices available in Braina.
Important: Using Offline voices on a slow processor may cause very high CPU usage and it may freeze/hang your computer especially if you use it along with other CPU intensive features like local LLM inference. If you have a slow processor, it is advisable to use a Lite/X_Lite offline voice model, fallback system voice or online voice.