LocalSTT VS STT-examples

There is already a way to piece it all together so that Kõnele, which normally needs a remote backend, just acts as a frontend for Vosk instead. I actually have this working on my phone, but, err, I think I'm forgetting one piece of the puzzle that needs downloading. Actually, I think you need to install a fork of this that supports English, unless you're fine with speaking Catalan. I forget which fork that is, however.

For an experimental Speechtotext, you can search localstt on github and use it with kõnele(F-Droid). And set it up as voice assistant in default apps settings. Not a great alternative, but something to fallback on. LocalSTT

You might want something like LocalSTT if it's on mobile: https://github.com/ccoreilly/LocalSTT

You're welcome! Your post actually got me a little interested in seeing what's new with Coqui STT (where all the old Mozilla STT folks moved to) and it seems someone was working on a WebAssembly binding for it, so one could probably finagle something themselves for testing purposes (the bandwidth of loading the model for every user on every page load is unfeasible from a production cost standpoint though)

I will add https://github.com/coqui-ai/STT, which is a continuation of DeepSpeech. Also, I've been messing around with https://github.com/ideasman42/nerd-dictation, which works on a VOSK backend - accuracy is decent, especially with the bigger model.

I don't know of a complete app, but https://github.com/coqui-ai/STT, which grew out of the now-unmaintained Mozilla Deepspeech project, works well and is easy to use. It could be a good starting point if you're comfortable writing a little code.

👉CoquiSTT - https://github.com/coqui-ai/STT

https://github.com/coqui-ai/STT-examples

If you have any more specific requirements then we can point you in the right direction. Or just join us on Matrix: https://app.element.io/#/room/#coqui-ai_STT:gitter.im :)