Speech Arena ranks top TTS (AI voice) models through crowd-sourced blind listening tests. Listen to two samples, pick the better voice, and help build a public leaderboard you can trust. Rankings use Elo with a dynamic K-factor (new voices adjust fast, established ones move slowly).
PoC suggesting that audio transformations* designed to make synthetic voices sound more credible are becoming obsolete in the face of spoofs generated by modern TTS/VC systems (2025). * Transformations proposed in the article “Breaking Security-Critical Voice Authentication”
Elevenlabs SRT Suite - Auto Translate .SRT files in different languages. Then creates a TTS with elevenlabs. The software then stores each sentence in a separate audio file that you can review and edit and replace any given sentence/audio-file before giving the final merge command.
This is a Text To Speech Testing setup to test your free and local tts models.
This project includes a Python script for fine-tuning a text-to-speech (TTS) model. The script utilizes custom datasets and use CUDA for accelerated training.
Assignment 2: Fine-tuning Text-to-Speech (TTS) Models for English Technical Speech and Regional Languages
Stop paying monthly fees. High-fidelity voice cloning on your own GPU. No limits, no credits.
Small TTS voice model for Apple Silicon
"A production-quality local Text-to-Speech (TTS) desktop studio. Run completely offline zero-shot voice cloning, and sub-second real-time streaming."
Pure C CPU inference engine for Irodori-TTS (Japanese TTS) with int8/VNNI paths — 4.5x faster.
AI SST model that takes all the noise from you classes and then gives it to you with time stamps and more!
Build text-to-speech applications with this curated guide for real-time agent streaming and high-fidelity offline synthesis.
A Streamlit web app for AI-powered voice cloning using Coqui XTTS v2. Record or upload reference voices, clone speech in multiple languages, and generate natural audio outputs.
🐱 Kitten TTS Studio using local onnx models TTS Offline
Audiobook Simplifier is a tool that creates audiobooks from text documents or eBooks using TTS (Text-to-Speech) technology.
Clone any voice from audio file using Qwen3 TTS to generate speech.
ReVoice — free, local-first voice cloning & audiobook studio: Qwen3-TTS cloning, chapter narration, M4B export. No cloud, runs on 4GB VRAM.
XTTS fine-tuning via CLI
To associate your repository with the tts-model topic, visit your repo's landing page and select "manage topics."