VITS TTS
The end-to-end TTS architecture that combines a variational autoencoder, normalizing flows, and adversarial training.
Wrap your text in SSML tags for precise control:
<speak><prosody rate="slow">Slow speech</prosody></speak>
Tags déi d'gewielt Modell verstinn - klickt fir eng an Ärem Text ze setzen wou se geschitt:
Dëse Modell liest einfache Text, sou datt Inline-Tags ignoréiert ginn. Fir Tag-baséiert Emotiounen, wielt e expressiven Modell wéi Orpheus oder Bark.
Eegen Aussproochen definéieren (Wuert = Aussprooch):
Iwwer 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.
Bescht fir: General-purpose text-to-speech with natural prosody
All sichen VITS StimmenOp ee Bléck
- Entwéckler
- Jaehyeon Kim et al.
- Lizenz
- MIT
- Tier
- free
- Geschwindegkeet
- fast
- Sprooche-Klonen
- Nee
- Sproochen
- English, German, Spanish, French, Portuguese, Dutch, Finnish, Hungarian, Bulgarian, Japanese, Polish
- Maximal Zeichen
- 2000