Does it uses device resources for comuputation or some cloud server ? about speech_to_text

[sowens-csd]

The package uses the platform dependent implementation of speech processing. So whether it depends on device or cloud will vary based on the particular device and the options you choose. Some mobile devices do support on-device processing so if you request it and it is supported the device will be used. However the package doesn't really dictate that since it is dependent on the details of the device implementation of the service.

[EuclidStellar]

so if it uses on-device processing then how it works? like I tried to use the whisper-medium model by OpenAI it was similar in comparison to speech_To_text in terms of confidence level and response time
so I want to know how this speech_to_text works ? is it trained on a model, or does it use some algorithms to calculate recognised words by pattern matching with phonemes?

please explain in simple steps what happen in between ( from when I speak to it get display in text field ) ?
I will be more than thankful to you for this

[sowens-csd]

The details vary by platform but the plugin does not handle speech recognition. Instead it calls platform APIs that provide speech services for that platform. So iOS has an implementation as does Android, etc.

[EuclidStellar]

Thanks