Google Launches Gemini 3.5 Audio Updates With New Transcribe Tool

Google has expanded its Gemini Audio lineup with new Gemini 3.5 models, rolling out enhancements just as users continue to anticipate the delayed launch of the Gemini 3.5 Pro model originally promised for June. The latest updates focus on improving voice-controlled artificial intelligence features and introducing advanced transcription capabilities.

The headline addition is Gemini 3.5 Transcribe, an entirely new model designed to convert speech into text with significantly higher precision. Google states that 3.5 Transcribe represents a major architectural advancement over its previous transcription model, Chirp 3, offering massive improvements in multilingual performance and lower wording error rates.

The new transcription model is packed with features aimed at making voice editing and documentation seamless for users:

• Automatic removal of verbal filler words such as "um" and "uh" to clean up text naturally.
• Speaker attribution supporting up to three distinct voices in pre-recorded audio, complete with word-level timestamps.
• Custom vocabulary integration, enabling the model to adapt to specialized jargon and unique spelling requirements automatically.

Alongside the Transcribe model, Google is launching Gemini 3.5 Live and Gemini 3.5 Live Experimental, which build upon the speech recognition technology behind Gemini's voice chat mode. Gemini 3.5 Live offers improved handling of mid-sentence interruptions and language recognition, while the Experimental version narrates its progress step by step in real time during complex reasoning tasks.