Voices and Emotions

Original Voice Soundboard ships with 54+ voices and 19 emotions out of the box. You can also describe the style you want in plain English.

Voices

Kokoro provides voices across multiple accents and genders. Voice names follow the pattern {accent}{gender}_{name}:

• af_ — American female
• am_ — American male
• bf_ — British female
• bm_ — British male

from voice_soundboard import VoiceEngine

engine = VoiceEngine()

# American female
result = engine.speak("Hi there!", voice="af_bella")

# British male
result = engine.speak("Cheerio!", voice="bm_george")

Voice presets

Presets bundle a voice, speed, and style into a single name for common use cases:

Preset	Voice	Speed	Style
assistant	af_bella	1.0	Friendly, conversational
narrator	bm_george	0.95	Calm, documentary
announcer	am_michael	1.1	Bold, energetic
storyteller	bf_emma	0.9	Expressive, varied
whisper	af_nicole	0.85	Soft, gentle

result = engine.speak("Breaking news!", preset="announcer")
result = engine.speak("Once upon a time...", preset="storyteller")

Emotions

19 emotions are available for fine-grained tonal control:

happy, sad, angry, excited, calm, fearful, surprised, disgusted, contemptuous, tender, proud, ashamed, guilty, anxious, nostalgic, hopeful, determined, confused, amused

result = engine.speak("I can't believe it!", emotion="surprised")
result = engine.speak("We will make it.", emotion="determined")
result = engine.speak("Those were the days.", emotion="nostalgic")

Natural language style

Instead of picking a preset or emotion by name, describe the style you want in plain English:

result = engine.speak("Good morning!", style="warmly and cheerfully")
result = engine.speak("The results are in.", style="serious, measured, like a news anchor")
result = engine.speak("I missed you.", style="soft and a little sad")

The engine maps your description to the closest combination of voice parameters.

Combining controls

You can layer voice, emotion, and style together:

result = engine.speak(
    "Welcome to the show!",
    voice="am_michael",
    emotion="excited",
    style="bold and energetic"
)

Presets serve as a convenient starting point, but direct voice + emotion + style gives you the most control.