Audio Tags for the Ekspresif Model

Ekspresif is the most expressive model in Possibl, with the most complete audio tag support: short instructions in square brackets, written right in your script, to control emotion, speaking style, and even sound effects, without changing the voice you picked.

How to use it

Write a tag in square brackets, right between the sentences of your script. A tag is not spoken. It only changes how the voice speaks around it. You can combine several tags in one sentence.

[whispers] Something is coming… [sighs] I can feel it.

Tag Categories

Emotion

[sad]Sad
[angry]Angry
[excited]Excited
[curious]Curious
[sarcastic]Sarcastic

Human reactions

[laughs]Laughs
[sighs]Sighs
[crying]Crying
[clears throat]Clears the throat
[gulps]Gulps

Speaking style & pace

[whispers]Whispers
[shouts]Shouts
[pause]Pause
[rushed]Rushed
[drawn out]Slowed down or drawn out

Accent & character

[strong French accent]A strong French accent (replace "French" as needed)
[pirate voice]A pirate-style voice

Sound effects

[applause]Applause
[explosion]Explosion
[gunshot]Gunshot

This is not a complete list. You can experiment with other tags as long as they describe a quality of the sound (tone, reaction, effect) and not a visual action.

Punctuation & Capital Letters

An ellipsis (…) adds a pause and emphasis.
CAPITAL LETTERS add emphasis to a word, for example "This is IMPORTANT to remember".
Ordinary punctuation (periods, commas) helps the rhythm of speech feel natural.

Tips for the Best Results

Tags are written in English exactly as shown. This is the vocabulary the model recognizes, not a free translation.
The voice you pick matters a lot. Some voices respond to tags more than others.
Do not stack too many tags on one line. The result can become unstable.
Some experimental tags (such as certain accents or sound effects) can give different results between voices.
Generate a short script first to hear the effect before processing a long one.

Try it in Possibl