MahaTTS: A Unified Framework for Multilingual Text-to-Speech Synthesis

Jaskaran Singh, Amartya Roy Chowdhury, Raghav Prabhakar, Varshul C. W
8/5/2025
eess.AScs.CL

Official code: dubverse-ai/MahaTTSv2 (3 stars, last commit 2026-02-04).

Abstract

Current Text-to-Speech models pose a multilingual challenge, where most of the models traditionally focus on English and European languages, thereby hurting the potential to provide access to information to many more people. To address this gap, we introduce MahaTTS-v2 a Multilingual Multi-speaker Text-To-Speech (TTS) system that has excellent multilingual expressive capabilities in Indic languages. The model has been trained on around 20K hours of data specifically focused on Indian languages. Our approach leverages Wav2Vec2.0 tokens for semantic extraction, and a Language Model (LM) for text-to-semantic modeling. Additionally, we have used a Conditional Flow Model (CFM) for semantics to melspectogram generation. The experimental results indicate the effectiveness of the proposed approach over other frameworks. Our code is available at https://github.com/dubverse-ai/MahaTTSv2

View on arXivView PDF

Code Implementations(1)

dubverse-ai/MahaTTSv2Official100%
30PythonJul 30, 20258 months ago

Similar Papers

MParrotTTS: Multilingual Multi-speaker Text to Speech Synthesis in Low Resource Setting

85% match2023

IndicVoices-R: Unlocking a Massive Multilingual Multi-speaker Speech Corpus for Scaling Indian TTS

84% match1 repo652024

RASMALAI: Resources for Adaptive Speech Modeling in Indian Languages with Accents and Intonations

84% match2025

A2TTS: TTS for Low Resource Indian Languages

83% match2025

Kinship in Speech: Leveraging Linguistic Relatedness for Zero-Shot TTS in Indian Languages

82% match2025

MunTTS: A Text-to-Speech System for Mundari

82% match1 repo82024

Cite this paper

@article{singh2025mahatts,
  title  = {MahaTTS: A Unified Framework for Multilingual Text-to-Speech Synthesis},
  author = {Jaskaran Singh and Amartya Roy Chowdhury and Raghav Prabhakar and Varshul C. W},
  year   = {2025},
  eprint = {2508.14049},
  archivePrefix = {arXiv},
  url    = {http://arxiv.org/abs/2508.14049v1}
}

Add a badge to your README

Shows whether official code is linked for this paper.

Code of Paper badge
[![Code of Paper](https://codeofpaper.com/badge/2508.14049.svg)](https://codeofpaper.com/paper/2508.14049)

Discussion