Local AI · macOS · iOS · iPadOS
AI that
stays home.
Mira runs entirely on your Mac, iPhone, and iPad — powered by Apple Silicon, no API keys, no subscriptions. Your conversations stay on your devices.
Native on every device
How it works
Local inference
Powered by Apple Silicon's Neural Engine via mlx-lm. Run Llama, Gemma, Qwen, and other open models — fully offline, zero API keys, no metered costs.
No cloud
Your conversations stay on your Apple devices. Sync between Mac and iPhone uses your personal iCloud — no Mira servers, no third-party data handling, ever.
macOS + iOS + iPadOS
Native SwiftUI apps sharing one backend. The Mac runs the model; your iPhone and iPad connect over your home network for a seamless multi-screen experience.
Open source
The inference backend (mira-core) is MIT licensed — inspect it, modify it, self-host it. Full transparency into what runs on your machine.
Ecosystem
Three repos,
one system.
mira-core
FastAPI backend with mlx-lm inference, RAG engine, web search, and SSE streaming. Runs on any Apple Silicon Mac.
Python · FastAPI · mlx-lm
mira-apps
Native SwiftUI client for macOS, iOS, and iPadOS. One codebase, three platforms.
Swift · SwiftUI · iOS · macOS
Vera
Companion iOS app in the Mira ecosystem, focused on document intelligence and code assistance.
Swift · SwiftUI · iOS