Text to speech programs (TTS)

[AndreaJayne]

I'm actually looking for a good one for just reading audiobooks on an android phone. Doesn't have to be perfect, but better than the default google ones would be lovely. Might any of these work?

[brandynette]

I'm actually looking for a good one for just reading audiobooks on an android phone. Doesn't have to be perfect, but better than the default google ones would be lovely. Might any of these work?

You can run coqui-tts locally & feed it your text

[Like Ra]

You can run coqui-tts locally

https://github.com/coqui-ai/TTS
https://docs.coqui.ai/en/latest/models/xtts.html

[brandynette]

you can also use my AIGF with a running coqui-tts set to default eng jenny
https://bambisleep.chat/api/tts?text=<BAMBI-TEXT-HERE>
if you open them like this it should give you a standalone audio player with 3 buttons to download your bambi audio.

just in case it still too complicated for thy fair lady of bambisleep ive added a video. might be a bit loud


2025-01-23 02-32-59.mp4 (Size: 37.9 MB )

[AndreaJayne]

Love you work, the voice is just lovely and I think this approach is great for pasting in scripts etc. My use would be to provide test for an audiobook in android. The book reading app handled the large text, remembers where you left off etc but you then select a TTS engine to facilitate the speaking. Love wait you have done here but not sure how to get it working on a phone,

[brandynette]

O.o
Yeah, a model who listens to the audio file, create a bambisleep themed, goal targeted test to see how proficient bambis are at being bambis!

IM IN!
if you handle the mobile app ill handle the desktop version & we joing bambi forces!

[brandynette]

Can anybody recommend good text to speech programs? Both standalone and browser-compatible (think of JavaScript reading from a file).

this is exactly what you are looking for. i warn you, it "speaks" analog, it dosent sound bad as its not a human voice. its literally just calculated waveforms using the audio api
https://www.npmjs.com/package/speak-tts

[oz88]

After some experiments with Melo TTS and Spark TTS, Kokoro TTS has proven to be the easiest regarding installation and the best regarding quality and stability.

For those searching for a shortcut to running Kokoro TTS locally:

Option (1) Install via Github

Powershell:
git clone https://github.com/PierrunoYT/Kokoro-TTS-Local
cd Kokoro-TTS-Local
py -3.10 -m venv venv
.\venv\Scripts\activate
pip install -r requirements.txt
python gradio_interface.py
open the browser link (usually http://localhost:7860)

Option (2) Install Portable Package incl. Windows GUI

https://github.com/mirbehnam/Kokoro-TTS-...s/releases

Select the package of your liking.

Btw. the voices seem to be identical with the ones used by https://elevenlabs.io/.

[AI_addict_hypnofan]

Be interested to hear otherwise but Elevenlabs is the best TTS at the moment I think.
It's definitely better than Google Native Audio TTS, though that may just be down to Elevenlabs having far superior voices.

This is their new model that just launched:

Source: https://www.youtube.com/watch?v=zv_IoWIO5Ek

[Like Ra]

Elevenlabs

Now with the link: https://elevenlabs.io/pricing