Skip to main content
Fish Audio S2.1 Pro and S2 Pro are expressive multilingual text-to-speech models that share the same voice library. Voices are referenced by a Fish Audio voice-model ID. Pass the voice_id from the tables below as the reference_id field in your request.
Leave reference_id unset to use the model’s default voice, or supply inline reference audio over MessagePack for one-shot voice cloning.

Supported languages

English

Spanish

Korean

Japanese

Arabic

Chinese

French

German

Turkish

Portuguese

Malay

Vietnamese

Dutch

Russian

Croatian

Czech

Danish

Finnish

Georgian

Greek

Hindi

Hungarian

Indonesian

Italian

Polish

Romanian

Swedish

Tagalog

Thai

Afrikaans

Albanian

Amharic

Armenian

Azerbaijani

Bengali

Bulgarian

Catalan

Estonian

Gujarati

Hebrew

Icelandic

Kannada

Kazakh

Latvian

Lithuanian

Malayalam

Marathi

Nepali

Norwegian

Persian

Punjabi

Sinhala

Slovak

Swahili

Tamil

Telugu

Ukrainian

Urdu

Xhosa