Google may soon allow customization of Gemini voice parameters

Google has released Gemini 3.5 Transcribe, an AI model designed for speech-to-text conversion that polishes voice input by removing filler words and allowing on-the-fly corrections. This model offers significant speed improvements and reduced error rates compared to its predecessor, Chirp 3, and will be integrated across the Google ecosystem.

Google announced Gemini 3.5 Transcribe, a new AI audio model that offers enhanced speech recognition, automatic language detection, and the ability to convert unstructured speech into formatted text. This development matters because it integrates advanced voice input capabilities across Google products and web environments, potentially changing how users interact with digital interfaces.

Google has launched Gemini 3.5 Transcribe, a new speech-to-text model available through the Gemini API, designed for intelligent voice interactions. This release allows developers to integrate advanced transcription capabilities, previously used in Google's own products, into their applications.

Google released Gemini 3.5 Transcribe, an updated AI transcription model that automatically removes filler words, formats text, and adapts to specialized jargon through custom vocabularies. This update improves multilingual performance and accuracy compared to its predecessor, Chirp 3, and is rolling out to macOS Gemini app users and Android's Rambler dictation feature.

Google introduced Gemini 3.5 Transcribe, a new speech-to-text model that offers improved precision and handles natural speech patterns, background noise, and jargon. This model is already integrated into products like Gboard Rambler and the Gemini macOS app, and will be coming to Chrome, enhancing voice interaction and transcription capabilities across Google's ecosystem.

Google's 'Rambler' voice-to-text feature on the Pixel 11, powered by Gemini Intelligence, effectively processes speech into coherent messages by removing filler words and adding formatting. However, it lacks a live transcription preview, which requires users to wait for full processing before seeing the output, a feature that Nothing's upcoming 'Essential Voice' update demonstrates.

Google has updated Gboard's Writing tools on the Pixel 11 series with Gemini Intelligence to include "Personalized suggestions." This feature provides draft suggestions tailored to a user's personal style based on current conversation, text input, and typing patterns, utilizing on-device and cloud-based Private AI Compute.

Google's new Gboard Rambler feature on the Pixel 11 series processes speech after recording, rather than providing real-time transcription. This allows it to refine text, filter filler words, and enable post-transcription editing and style changes, differing from previous real-time speech-to-text systems.

Google's new Pixel 11 series debuts several AI-powered features, including Gboard Rambler for refined voice transcription, an updated Proactive Assistance (formerly Magic Cue) with expanded contextual suggestions, and a location-aware At a Glance display. These updates integrate Gemini Intelligence to automate tasks and provide more relevant information directly on the device.

Google is rolling out advanced voice control capabilities for Gemini on macOS, including intelligent dictation that refines spoken words and a feature that understands screen context to perform complex tasks. This update allows users to interact with their desktop using natural language for tasks like summarizing documents, rewriting text, and generating images, aiming to provide a speech-to-text experience similar to Gboard Rambler.

Google's upcoming Gemini update could enable users to customize voice characteristics such as Energy, Formality, Speed, and Warmth. This feature will enhance user interaction with Gemini across various applications, likely improving personalization options for voice interactions.