Hearthscale

Hearthscale lets you run and build with local models for code, voice, image, video, music, and transcription, all on your own device. Key features: • Meeting notes generated from recordings • System-wide voice dictation • Image and music generation studios • A built-in agent • Marketplace apps you can install in one click • End-to-end encrypted sharing with your other devices. Hearthscale is free, private, and requires no account. Hearthscale Server can be deployed in one click or with Docker on your most capable machine, then connects to the desktop client automatically to serve local models to every device you own. Your prompts, recordings, and files stay on hardware you control. Create songs from a few words with models like MiniMax Music 3, ACE-Step 1.5, and Stable Audio 3, or turn meeting audio into organized notes and action items. Perfect for writers, musicians, students, small business owners, and privacy-conscious professionals who want powerful tools without subscriptions or cloud lock-in. Developers can also build custom apps on top of open local models.