Expanded Gemini’s enterprise offerings with Gemini 3.5 Transcribe, supporting more than 85 languages and improved handling of specialized terminology.

Google introduced Gemini 3.5 Transcribe, expanding its enterprise voice-AI portfolio through the Gemini Enterprise Agent Platform and Gemini API public preview. The model automatically detects more than 85 languages, handles accents and code-switching, removes filler words, formats speech, and supports custom vocabularies for jargon and proper names. Its live version offers sub-second streaming, while batch transcription adds word-level timestamps and speaker identification for up to three people—potentially reducing manual editing in contact centers, meetings, and specialized workflows.