Use your iPhone to clone your voice*, chat with your documents and create AI personas. All processing is done privately on your phone. The app's source code is available on GitHub.
*Voice cloning uses our server once per voice, then works offline
Clone your voice in minutes. Use it on-device in real-time, 1 second of voice generated per second on modern iPhones.
Chat with your documents (PDFs and Markdowns) using on-device RAG (retrieval-augmented generation), your documents never leave your iPhone.
Make your virtual personas (characters) based on your previous conversation history. All processing is done privately on-device.
On-device speech-to-text and text-to-speech. Talk naturally, no internet required.
Using Apple Foundation Models (on-device) or lightweight Llama/Qwen/Gemma models (one-time download, works offline).
Local Voice is my personal project, all client code is available on GitHub. iOS/iPadOS 26+, XCode 26+.
View on GitHubOn-Device Processing: Local Voice processes your text, documents, and AI on-device. Your content never leaves your device.
Voice Creation: Audio samples are securely sent to our server for one-time processing and immediately deleted.
Diagnostics: Anonymous crash reports and performance data may be collected to improve the app. No personal information is included.