Google has officially announced Gemini 3.5 Transcribe, a new AI model designed to streamline speech-to-text processes. This new AI model is designed to transcribe voice inputs into text faster and more accurately. This technology is currently actively used in the Gboard Rambler feature found in Pixel 11 devices. Google announced that this model will be rolled out across its entire ecosystem in the coming period.

Fast and Smooth Transcription with Gemini 3.5 Transcribe

Compared to Google’s previous speech-to-text engine, Chirp 3, a significant increase in performance is observed with the new model. Gemini 3.5 Transcribe offers approximately 70% faster speed in the process of converting audio data to precise text.

This technology not only hears the words accurately but also analyzes what the user means. Unnecessary terms such as “uh” or “um” that are frequently used during speech are automatically removed by the model.

In addition, instant edits can be made to the text if users correct themselves. For users who need special terminology, the system also takes into account the defined special vocabulary.

The system provides a more professional output by editing the text without disrupting the user’s speech flow. This provides great convenience, especially for users who need to take notes quickly.

Wide Language Support and Application Areas

Gemini 3.5 Transcribe supports 85 different languages ​​​​worldwide, appealing to a wide user base. It also has the ability to distinguish three different speakers simultaneously in previously recorded audio files.

These cleaning features offered by the model are quite successful in short blocks of text. However, the fact that artificial intelligence technically alters spoken words may not be suitable for some situations.

Google states that this model aims to take the voice input experience to a more professional level. Typo and hesitations experienced by users giving when voice commands will now be less of a problem.

In this period where voice input is gaining increasing importance in the technology world, Google’s initiative is seen as a valuable step. These improvements offered by the software can speed up the voice note-taking process in daily use.

This new AI-powered model is expected to increase efficiency, especially in editing long meeting notes or audio reports. Google aims to strengthen the role of voice input in text-based communication with this technology.

Do you think AI-edited texts distort the naturalness of original speech?

Google announced the new Gemini 3.5 Transcribe model, which speeds up speech-to-text conversion processes and eliminates errors.