Google DeepMind has launched Gemini 3.5 Transcribe, a new speech-to-text model designed for enhanced voice interactions. This model offers improved accuracy, handling background noise, jargon, and self-corrections, while also removing filler words and auto-formatting text. Developers can now integrate Gemini 3.5 Transcribe into their applications via the Gemini API, Google AI Studio, and the Gemini Enterprise Agent Platform, with options for real-time streaming or pre-recorded audio processing. AI
IMPACT Enhances voice interaction capabilities for consumers and developers, potentially accelerating adoption of voice-based AI applications.
RANK_REASON Frontier-lab model release with system card [lever_c_demoted from frontier_release: ic=1 ai=1.0]
- android
- Artificial Analysis
- Chirp 3
- Gemini 3.5 Transcribe
- gemini-3.5-transcribe-live
- Gemini API
- Gemini app
- Gemini Enterprise Agent Platform
- Google AI Studio
- Google DeepMind
- macOS
AI-generated summary · Google Gemini · from 1 sources. How we write summaries →