I was testing Essential Space on my Nothing Phone (3a) and noticed something interesting.
When I turn ON airplane mode and record a voice note, it shows:
“Using Whisper in Essential Space”
When I turn internet back ON, it shows:
“Using Gemini 2.5 Flash in Essential Space”
So I’m trying to understand the actual pipeline:
Is Whisper running locally on-device for speech-to-text, and then Gemini is used only when online for processing/summarization?
Or is Whisper also cloud-based and just labeled differently?
Would love if someone could clarify how this actually works internally (local vs cloud processing).