#speech-to-text

Gemini 3.5 Transcribe: Google's New Speech-to-Text Model That Actually Cleans Up After You

Google shipped a transcription model that removes filler words, handles self-corrections, and streams with sub-second latency. Here's what it means for developers.