Skip to content

Voices & Presets

Language is auto-inferred from the voice ID prefix — no configuration required.

PrefixLanguage
af_ / am_English (American)
bf_ / bm_English (British)
jf_ / jm_Japanese
zf_ / zm_Mandarin Chinese
ef_ / em_Spanish
ff_French
hf_ / hm_Hindi
if_ / im_Italian
pf_ / pm_Brazilian Portuguese

Five curated presets for common use cases:

PresetVoiceSpeedDescription
narratorbm_george0.95Calm, clear, documentary style
announceram_eric1.1Bold, energetic, broadcast style
whisperaf_sky0.85Soft, intimate, gentle
storytellerbf_emma0.90Expressive, varied pacing
assistantaf_jessica1.0Friendly, helpful, conversational

Six humor presets are also available for sensor-humor integration. Use the mood parameter on voice_speak with one of: dry, roast, chaotic, cheeky, cynic, zoomer.

English (American) — 14 voices including Aoede (musical), Bella (warm), Heart (caring), Jessica (professional), Eric (confident), Fenrir (powerful), Puck (playful).

English (British) — 6 voices including Alice (proper), Emma (refined), Fable (storytelling), George (authoritative).

Japanese — 5 voices including Alpha (clear), Gongitsune (storytelling), Nezuko (gentle).

Mandarin Chinese — 8 voices including Xiaobei (bright), Yunjian (authoritative), Yunxi (friendly).

Spanish — 3 voices. French — 1 voice. Hindi — 4 voices. Italian — 2 voices. Brazilian Portuguese — 3 voices.

Use voice_status to get the full list of available voices at runtime.