Skip to content

LiveSTT

Offline speech-to-text and live translation

LiveSTT uses on-device AI to transcribe speech and translate it in real time, with floating subtitles that work across videos, podcasts, meetings, lectures, and everyday conversations.

Preview artwork for LiveSTT
01

Watch foreign-language videos and creator content with live subtitles

02

Turn podcasts, classes, lectures, and meetings into readable text

03

Understand travel announcements and multilingual conversations on the go

Always-on-top floating subtitle window

On-device speech recognition and offline translation

Microphone and internal audio transcription

Does LiveSTT require an active internet connection to transcribe speech?
No. LiveSTT processes 100% of speech-to-text offline on-device, leveraging local neural engine weights. Transcriptions and live translations are calculated entirely locally, making it highly secure and reliable even in zero-connectivity environments like airplanes or remote locations.
What powers the high transcription accuracy of LiveSTT?
LiveSTT integrates advanced Apple Speech Analyzer frameworks under iOS 26 alongside local SenseVoice-Small models. This hybrid local execution achieves up to 98% accuracy on clear speech and operates at up to 3x real-time speeds, providing swift and precise transcriptions.