Has anyone been able to solve this? Or is this impossible until native plugins are open?
Users want to talk to fill out inputs. I tried an external API call but it’s too slow. Even the fastest models are too slow.
React uses OS-native speech-to-text APIs, which are:
-
Faster
-
Zero latency
-
100% offline on most phones
-
Free
-
Ultra accurate for dictation
-
Designed exactly for “tap mic → fill input”
This is how every serious app does it.