MoeGoe

MoeGoe is a free, open-source graphical front end for VITS text-to-speech inference. Released under the MIT license and downloaded as a ready-to-run executable, the project has collected over 2,400 GitHub stars; its last code updates landed in 2023, but it remains one of the simplest ways to run VITS voice models on a desktop.

Overview

VITS is an end-to-end neural text-to-speech model family popular for anime-style and character voices, but running it normally means setting up a Python environment and writing inference code. MoeGoe removes that friction: you load a trained VITS model and its config, type or load your text, pick a speaker, and generate speech through a simple window. It was built primarily for the Japanese and Chinese VITS model community, where fans train voices for virtual characters and share the resulting models.

Key Features

• Point-and-click GUI for VITS inference, no coding or Python setup needed
• Loads community-trained VITS models and their configuration files
• Multi-speaker selection for models that contain several voices
• Adjustable speech parameters such as noise scale and length scale
• Supports Japanese and Chinese text input used by most VITS models
• Distributed as a standalone executable for Windows users

💳 Pricing

MoeGoe is completely free and open source under the MIT license. You only need a VITS model to run, and many community models are shared at no cost.

Who Is It For

MoeGoe is for anime and VTuber fans, doujin creators and hobbyists who want to synthesize character speech with VITS models without touching a command line or training pipeline.