VITS

VITS TTS

The end-to-end TTS architecture that combines a variational autoencoder, normalizing flows, and adversarial training.

Whakawhanake mō te tepe o ngā tohu 5,000

Whāriki i tōna kupu i roto i ngā tohu SSML mō te whakahaere tika:

<speak><prosody rate="slow">Slow speech</prosody></speak>

E mōhio ana ngā tohu ki te tauira i kōwhiria - ka kōwhiria kia whakawātea tētahi ki roto i tōna kupu i reira ka puta ai:

Ka pānui tēnei tauira i te kupu noa, nā reira ka whakakāhoretia ngā tohu ā-waitara. Mō te āhua o te tohu-taihi, ka huri ki tētahi tauira whakamārama pēnei i a Orpheus, Bark rānei.

Ka tautuhia ngā tohutohu ā-ringa (wāhi = tohutohu):

-12 +12
0.5x 2.0x
Waihoki me Piper, VITS, MeloTTS
Ka puta tēnei te oro i waihangatia e koe. Ka kōwhiria tētahi tauira, ka tāurua te kupu, a, ka kōwhiria te Whakatū.
Kua angitu te whakaputanga oro
0:00
Waihoki i te oro Whakahua.srt Ka ngaro te pānga i roto i te 24h
Tauwhāinga wātea: te whakamahinga whaiaro. Whakawhiwhinga hokohoko mai i te $5/mo
E manakohia ana e TTS.ai? Whakapāpāho ki ōna hoa!

Mo VITS

VITS — Variational Inference with adversarial learning for end-to-end Text-to-Speech — was introduced by Jaehyeon Kim and collaborators in 2021 and became a foundational architecture for modern neural speech. Rather than the older two-stage pipeline, it synthesizes audio in a single parallel end-to-end pass, pairing a variational autoencoder with normalizing flows and a GAN-style adversarial training process to lift naturalness. At about 25M parameters and trained on ~585 hours, it produces natural prosody at fast inference speeds and supports multiple speakers. It serves as a solid general-purpose, free baseline and underpins many later models such as Piper and MeloTTS.

Pai mo: General-purpose text-to-speech with natural prosody

Ka tirohia katoa VITS ngā oro

I te tirohanga

Ka whakawhanakehia
Jaehyeon Kim et al.
Ka taea te whakawātea
MIT
Karaka
free
Āhuatanga
fast
Whakakōrero reo
Kāore
reo
English, German, Spanish, French, Portuguese, Dutch, Finnish, Hungarian, Bulgarian, Japanese, Polish
Kāri nui rawa
2000

VITS ngā oro

CSS10 (Dutch)

Dutch
Waihoki Neutral

CSS10 (Finnish)

Finnish
Waihoki Neutral

CSS10 (French)

French
Waihoki Neutral

CSS10 (German)

German
Waihoki Neutral

CSS10 (Hungarian)

Hungarian
Waihoki Neutral

CSS10 (Spanish)

Spanish
Waihoki Neutral

Common Voice (Bulgarian)

Bulgarian
Waihoki Neutral

Common Voice (Portuguese)

Portuguese
Waihoki Neutral

Default

English
Waihoki Neutral

MAI (Polish)

Polish
Waihoki Female

MAI (Ukrainian)

Ukrainian
Waihoki Neutral

VITS TTS - FAQ

VITS means Variational Inference with adversarial learning for end-to-end Text-to-Speech. It generates audio in a single parallel pass using a variational autoencoder, normalizing flows, and adversarial (GAN) training, rather than a two-stage pipeline.

Yes. VITS is MIT-licensed and in the free tier, so it can be used commercially.

On TTS.ai, VITS covers 11 languages including English, German, Spanish, French, Portuguese, Dutch, Finnish, Hungarian, Bulgarian, Japanese, and Polish, with multi-speaker support. It does not do voice cloning.
← Ko nga oro katoa