Podcawise TTS Models

ONNX model files for Chatterbox Multilingual TTS engine used by Podcawise - Desktop AI Podcast Studio.

Supported Languages (23)

Arabic, Danish, German, Greek, English, Spanish, Finnish, French, Hebrew, Hindi, Italian, Japanese, Korean, Malay, Dutch, Norwegian, Polish, Portuguese, Russian, Swedish, Swahili, Turkish, Chinese

Files

This repository hosts the following model files in Releases:

v1.1.0 - Multilingual (Recommended)

• speech_encoder.onnx / speech_encoder.onnx_data
• embed_tokens.onnx / embed_tokens.onnx_data
• language_model_q4.onnx / language_model_q4.onnx_data (INT4 quantized)
• conditional_decoder.onnx / conditional_decoder.onnx_data
• tokenizer.json
• Cangjie5_TC.json (Chinese character mapping)

Total download size: ~1.5GB

v1.0.0 - English Only (Legacy)

English-only version for smaller footprint if multilingual support is not needed.

Source

Models are sourced from onnx-community/chatterbox-multilingual-ONNX on Hugging Face.

License

The models are distributed under their original license terms.