Introducing Gemini 3.1 Flash TTS: A New Era in Text-to-Speech Technology
Google has unveiled a groundbreaking text-to-speech AI model called Gemini 3.1 Flash TTS. This new model is designed to offer enhanced controllability, expressivity, and overall quality. According to Google, it represents the most natural and expressive text-to-speech model they have developed to date.
In the Artificial Analysis TTS leadboard, a benchmark that gathers thousands of human preferences, Gemini 3.1 Flash TTS achieved an impressive Elo score of 1,211. The model has also been positioned in what Google refers to as the ‘most attractive quadrant’ on this leaderboard. This placement highlights the model’s ability to deliver strong performance while maintaining a low cost, making it a compelling choice for developers and businesses alike.
Key Features of Gemini 3.1 Flash TTS
One of the most significant improvements in Gemini 3.1 Flash TTS is its enhanced speech controllability. Users can now guide how the AI speaks by using natural language instructions. The model also introduces audio tags, which allow for more precise adjustments to vocal delivery. This includes control over speaking speed, pace, and overall delivery. As Google explains, “By embedding natural language commands directly into the text input, you can steer AI-speech output with improved levels of granularity.”
Another notable feature is the support for multi-speaker dialogue. Developers can create distinct characters with unique audio profiles, making it easier to generate realistic conversations. Additionally, Gemini 3.1 Flash TTS supports more than 70 languages, ensuring broad accessibility and usability across different markets. Google states, “Gemini 3.1 Flash TTS delivers high-fidelity speech and more precise control across more than 70 languages. These core optimisations bring advanced style, pacing and accent control to major markets.”
Additional Considerations
All audio generated by Gemini 3.1 Flash TTS includes a SynthID watermark. This invisible watermark is embedded in the audio and serves as a tool for detecting AI-generated content. This feature adds an extra layer of transparency and accountability, particularly important in today’s digital landscape where AI-generated media is becoming increasingly common.
Availability and Access
Developers interested in trying out Gemini 3.1 Flash TTS can access it in preview through the Gemini API and Google AI Studio. Enterprise users have the option to use the model in preview via Vertex AI, while Workspace users can access the new model through Google Vids. This wide range of access points ensures that both individual developers and large organisations can benefit from the advancements offered by this new text-to-speech model.
With its impressive capabilities and broad availability, Gemini 3.1 Flash TTS is set to redefine the possibilities of text-to-speech technology. Whether you’re looking to enhance user experiences, develop interactive applications, or explore new creative avenues, this model offers a powerful and flexible solution.
