OmniVoice Voice Design Guide: Accents, Attributes, and Tags
OmniVoice voice design builds a new voice from a few structured attributes, with no reference audio to clone. This guide covers every attribute, the ten English accents that make OmniVoice stand out, its thirteen paralinguistic tags, and where emotion in OmniVoice actually comes from.
Key takeaways
- OmniVoice designs a voice from five attributes: gender, age, pitch, style and accent. Each one has an Auto option.
- Accents are the standout: ten English accents, set straight from a menu.
- Thirteen inline tags such as
[laughter]and[sigh]add non-verbal sounds. Use one per sentence at most.- OmniVoice has no emotion control. Emotion carries over from a cloned reference clip; for selectable emotions, use Qwen3-TTS.
- Designed voices are most stable in English and Chinese.
Hear the OmniVoice Accents
Accents are what OmniVoice does best, so they are the quickest way to hear what voice design gives you. Each voice below was designed from scratch by setting OmniVoice's attributes, with no reference audio. Every one speaks English; the accent attribute does the work.
OmniVoice vs Qwen3-TTS: Structured vs Free-Form Voice Design
Both models design a voice from scratch, but they take opposite approaches. OmniVoice is better for accents and repeatable results; Qwen3-TTS is better for nuance described in your own words.
- OmniVoice input: structured attributes, picked from fixed options
- Qwen3-TTS input: a free-form text description
- Accents: OmniVoice offers ten English accents as direct options; in Qwen3-TTS you describe them in free text, which is less reliable
- Pick OmniVoice when: you need a specific accent or a result you can reproduce exactly, especially for English or Chinese speech
- Pick Qwen3-TTS when: you want something dramatic and characterful, like a game character, a villain or "a gravelly, world-weary detective"
Both models are in Voice Creator Pro, so you can switch based on the job. Our Qwen3 TTS prompting guide shows how to write those free-form descriptions, with examples.
The Five OmniVoice Voice Design Attributes
OmniVoice voice design is built from five attributes, and every one has an Auto option that hands the decision to the model. Set the attributes you care about and leave the rest on Auto.
| Attribute | Options |
|---|---|
| Gender | Auto, Male, Female |
| Age | Auto, Child, Teenager, Young adult, Middle-aged, Elderly |
| Pitch | Auto, Very low, Low, Moderate, High, Very high |
| Style | Auto, Whisper |
| Accent | Auto, plus ten English accents (listed below) |
Pitch is independent of gender and age. You can push a young adult voice lower or lift an elderly voice higher to fine-tune the character.
Style is currently focused. Use Auto for normal delivery, or Whisper for a soft, breathy read. Whisper suits ASMR, intimate narration, suspense and meditation content.
The OmniVoice model page covers the model beyond voice design.
Ten English Accents
OmniVoice offers ten English accents, plus Auto: American, British, Australian, Canadian, Indian, Chinese, Korean, Japanese, Portuguese and Russian.
Each one generates English speech in the chosen accent. You can voice an Indian-accented narrator, a British storyteller or a Russian-accented character without finding and cloning a real speaker.
OmniVoice Voice Design Recipes
A handful of attribute choices defines a distinct voice. Set the ones that matter, leave the rest on Auto, and use these as starting points:
- Warm British audiobook narrator: Female, Middle-aged, Moderate pitch, British accent
- Energetic American podcaster: Male, Young adult, Moderate pitch, American accent
- Gentle ASMR reader: Female, Young adult, Low pitch, Whisper style
- Elderly storyteller: Male, Elderly, Low pitch
- Young Indian-accented explainer voice: Female, Young adult, Moderate pitch, Indian accent
- Suspense narrator: Male, Middle-aged, Very low pitch, Whisper style, American accent
Anything not listed stays on Auto. Change one attribute at a time and regenerate, so you can hear exactly what each control does.
Set one of these recipes in OmniVoice and hear the voice in seconds
Try OmniVoice freeOmniVoice Paralinguistic Tags: The Full List
Paralinguistic tags add non-verbal sounds to the speech: the laugh, the sigh, the questioning "hm?" before a sentence turns. You type them inline in square brackets, and OmniVoice performs the sound at that exact point.
In Voice Creator Pro they live behind the smiley icon in the text input, so you can insert one at the cursor. Expression tags work on OmniVoice and Chatterbox only, so they do nothing on the other models.
OmniVoice supports thirteen tags:
[laughter]: a laugh. The most useful of the set, and the most natural sounding.[sigh]: an audible exhale. Fatigue, relief or resignation, depending on the line around it.[confirmation-en]: a short affirmative "mm". Agreement or acknowledgement mid-conversation.[question-en]: a questioning "hm?" Puzzlement or a request to repeat.[question-ah]: a questioning "ah?" Slightly more open and surprised than[question-en].[question-oh]: a questioning "oh?" Reads as dawning realization or mild doubt.[question-ei]: a questioning "ei?" A sharper, more startled query.[question-yi]: a questioning "yi?" Curiosity at something unexpected.[surprise-ah]: a surprised "ah!" The general-purpose surprise sound.[surprise-oh]: a surprised "oh!" Closer to recognition than shock.[surprise-wa]: a surprised "wa!" Bigger and more delighted, good for awe or excitement.[surprise-yo]: a surprised "yo!" A short, punchy exclamation.[dissatisfaction-hnn]: a displeased "hnn". A grumble, for annoyance or reluctance.
Read the suffix as the sound, not as a language. The part after the hyphen is the vowel the model voices, which is why there are five question tags. [question-en] is not the English variant of anything, so pick by the sound you want to hear.
How to Use OmniVoice Tags in a Script
Drop a tag straight into the line, before the words it colours or after them:
[laughter] You really got me there.
So I was walking down the street [laughter] and you won't believe what happened next [surprise-oh]
[question-en] Are you sure that's what she said?
- One per sentence, at most. Tags read as punctuation for the voice. Stacked together they sound like a soundboard.
- Position matters. A tag at the start colours the delivery that follows. A tag at the end reads as a reaction to what was just said.
- Generate and listen. The five question tags are close cousins, and the right one is easier to hear than to predict. Swap one for another and regenerate.
- They are sounds, not directions. A tag makes the voice produce a specific noise. It does not tell the model to "sound sad" across a paragraph.
Where Emotion Comes From in OmniVoice
OmniVoice has no emotion control. There is no emotion menu, slider or tag. The tags above produce non-verbal sounds, and the Style attribute offers only Auto and Whisper.
Instead, OmniVoice carries the emotion of the reference audio you cloned from. Clone a shaky, upset voice and the output stays shaky; clone a bright, energetic read and the speech keeps that energy. For a designed voice with no reference audio, the delivery follows the attributes and the text itself.
To select an emotion directly, use Qwen3-TTS, which has thirteen emotions at five intensity levels each. It also takes free-form descriptions of how a line should be delivered.
Both models are in Voice Creator Pro, so pick per job: OmniVoice for accents, speed and carrying a reference's feeling, Qwen3-TTS to dial an emotion in explicitly. The voice prompting guide covers Qwen3 descriptions, and how to add emotion to text to speech applies emotions across a longer piece.
Tips for Better OmniVoice Voice Design
Start with Auto, then narrow. Generate once with most attributes on Auto to hear the model's default, then lock in the ones you want to change. This is faster than setting all five blind.
Pitch is your fine-tuning dial. Once gender, age and accent give you the right character, use pitch to nudge it warmer (lower) or brighter (higher).
Lean on accents for character variety. For a cast of several voices, accent is the fastest way to make them sound like different people, even with similar gender and age settings.
Change one thing at a time. Adjusting a single attribute per generation shows what each control contributes, so you can dial in the voice you hear in your head.
Voice design is most stable in English and Chinese. OmniVoice's voice design was trained on English and Chinese data. You can still synthesize other languages, but expect cloning to be the steadier option outside those two.
Voice Design vs Cloning vs Prompting
These three get conflated constantly, so here is the difference:
- Voice design (this guide): build a new voice from attributes. No recording needed.
- Voice cloning: copy a specific real voice from a short reference clip (3 to 10 seconds). Use this when you want a particular person's voice.
- Voice prompting: control the delivery, emotion and performance of a voice through text instructions (for example, our DramaBox prompting guide). This shapes how a voice acts, not who it is.
Design Your Voice in Voice Creator Pro
Voice Creator Pro includes OmniVoice voice design with every attribute above (gender, age, pitch, style and all ten accents) in a point-and-click interface. There is no setup or prompt engineering required.
Designed and cloned voices both work inside projects, so you can use them for full audiobooks, YouTube voiceovers and other long-form narration. Projects support multiple speakers, so you can build a designed cast and run an entire dialogue or chapter in one place.
The desktop app runs OmniVoice locally and offline, with unlimited generations and no subscription, on Windows and Mac. Or try it free in your browser, with no GPU or install required.
Try OmniVoice for free
Also available on Windows and macOS. One-time purchase, unlimited generations.