Applio

Applio is a local AI voice conversion suite for musicians making AI covers, streamers changing their voice live, and creators working with speech. It converts recordings or microphone input into another voice using community models or models you train yourself. Its software uses the MIT license.

The software runs on Windows, macOS and Linux, with a standalone desktop app and a Gradio interface you use in your browser. Local use doesn't require an account, and Applio collects no telemetry. Audio processing and model training run on your machine. For cloud use, it also offers a playground and Google Colab access.

Voice conversion includes pitch adjustment, audio cleaning and upscaling, plus batch processing for multiple files. Live conversion uses low-latency inference for singing, streams and calls. Text-to-speech lets you generate spoken audio and pass it through a voice model, so you can work from written material as well as recordings.

Custom model training starts with your own audio dataset and includes preparation, feature extraction, training and export. An audio analyzer checks dataset quality before training. If you want to combine voices, the voice blender merges multiple models and lets you control each model's contribution. Plugins support further customization, and the MIT license permits modification and redistribution for personal, research and commercial work. The project now focuses on security patches, dependency updates and occasional feature improvements, with less frequent releases.