Real time speech-to-text

Hello everyone!

I’m trying to build an application that performs speech-to-text in real time. I know I can use Whisper to achieve this with just a few lines of code. , but i’m still really confortable with bubble and I’m not sure how to integrate the two. I understand how to use Whisper through the Bubble API connector, but it seems that it only accepts MP3 files, which doesn’t allow for real-time processing.

Thank you so much in advance to everyone who can help!

Matilde

Hi @matilde

I have an AI BOT plugin that will do this for you, it utilises Deepgram’s new Aura feature and their speech to text (STT) and text to speech (TTS) services, amongst other things, but this may not be suitable if you’re not planning to use AI or anything like that.

@redvivi creates some good plugins, which may be better suited, have you seen this? Real-Time Speech To Text Plugin Demo (bubbleapps.io)

Plugin page: Real-Time Speech To Text Plugin | Bubble

Paul

1 Like

Other Realtime options are available https://forum.bubble.io/t/realtime-voice-transcription-with-assembly-ai-deepgram/249198

1 Like

thank you!!!
but i was wondering if it is possible to do it without the use of plugin.
is it possible to do it using the bubble connector and -for example - the whisper api?

I’m afraid not - Realtime uses websockets as the way of connecting to a service - the Bubble API connector is using HTTP calls. Which is different.