Expressive audio tags
Special codes in text that control how the voice sounds, like emotion and tone.
Expressive audio tags are like special instructions you can add to text before it's turned into speech. These tags tell the text-to-speech (TTS) system how to deliver the words, influencing the emotion, tone, and rhythm.
For example, you could add a tag like `<emotion type='sad'>` before a sentence to make the AI voice sound mournful, or `<speed rate='fast'>` to make it speak more quickly. This helps make AI-generated voices sound more natural and engaging, similar to how human speakers vary their delivery.
Expressive audio tags are control codes embedded within text input for text-to-speech (TTS) systems. These tags allow for fine-grained manipulation of prosody, including pitch, duration, loudness, and emotional affect, to enhance the naturalness and expressiveness of synthesized speech. They function by signaling the TTS engine to alter its acoustic modeling parameters or apply specific vocal styles based on the tag's content and context.
Google Gemini is previewing its 3.1 Flash Text-to-Speech model, highlighting expressive audio tags. These tags allow for finer control over the prosody and emotional delivery of synthesized speech, moving beyond basic voice generation.
- Gemini 3.1 Flash TTSGoogle · Gemini Models ·
- Pré-lançamento do Gemini 3.1 Flash TTS (conversão de texto em voz) | Gemini APIGoogle · Gemini Models ·
- Google Gemini
- Google Gemini's Gemini 3.1 Flash TTS previews the use of expressive audio tags to give users more control over the tone and emotion of generated speech.
- SSML
- Speech Synthesis Markup Language (SSML) is a broader standard that include
- Phonemes
- Phonemes are the basic units of sound in a language; expressive audio
- Embeddings
- Embeddings are numerical representations of text or audio; expressiv
The concept of using markup to control speech synthesis has roots in early TTS systems and evolved with standards like SSML. Specific 'expressive audio tags' are a more recent development in advanced TTS models aiming for richer emotional and stylistic control.
Experiment with a TTS tool that supports custom tags. Try inputting text with simple tags like `<emotion type='happy'>Hello there!</emotion>` and then `<emotion type='sad'>Hello there!</emotion>` to hear the difference in the voice's delivery. Observe how the AI's tone and inflection change based on the tag.
The Playbook: reusable how-tos, prompts, configs and tips pulled out of the news.
- PromptRole-play a student struggling with math for AI tutor simulationSimulate a student's math struggles and uncertainty for AI tutor evaluation. · 5 min
- TemplateReusable workflow templates for healthcare teamsTeams can create and share templates for common workflows like documentation or prior authorizations. · 15 min
- TemplateTemplate for building an AI tool tailored to your workA structured approach to defining and building a custom AI tool for specific business needs. · 15 min
- TemplateEvals CSV prompt list for agent skillsA CSV file template for listing prompts used in evaluating agent skills. · 20 min
- HeuristicDouble coding agent task completion with docs linksProviding coding agents with a documentation link significantly increases their task completion rate. · 2 min
Questions people ask
- Can I use these tags with any voice?
- Support for expressive audio tags varies by TTS model and voice. Some voices are trained to respond to a wider range of emotional and stylistic tags than others.
- Are expressive audio tags the same as SSML?
- Expressive audio tags are often a part of or similar to SSML, but the term specifically emphasizes the control over emotional and stylistic aspects of speech, beyond basic pronunciation or rate.
First seen on Subvolts Sep 3, 2026. Updated Sep 21, 2026. Written by an AI model from the items above and reviewed against them weekly; report an error.