BERT-VITS Series

The BERT-VITS Series is an open-source family of text-to-speech (TTS) models combining a multilingual BERT language backbone with the VITS architecture for expressive and natural-sounding speech generation.

• Multilingual support for high-quality, natural prosody
• Integrates BERT for improved linguistic context
• Supports emotional and stylistic speech variations
• Fine-tuning capability on custom datasets
• Available under a strong open-source AGPL-3.0 license

CLICK HERE TO DISCOVER BERT-VITS