AniSora Logo

High-fidelity, Open-source TTS: scalable multi-speaker synthesis with production-ready vocoders

Try live synthesis in the Dashboard demo.

Index‑TTS Demo

100% Open Source

Index‑TTS is fully open-source. Explore the code, models and demos — contribute or run locally.

Index‑TTS

Open-source TTS for research & production

The Index‑TTS project provides modular TTS components, model implementations, and examples to reproduce high-fidelity speech synthesis.

Multi-speakerHigh-fidelityProduction-ready
View on GitHub

Reproducible & extensible

Example training recipes, inference scripts and model weights are provided to help you reproduce results and extend the codebase.

Training recipes
Guides and configs to reproduce model training.
Inference examples
Prebuilt scripts for synthesis and evaluation.