Mobile
Smritsavant’s mobile apps run the same core as the desktop: the same local database, the same on-device embeddings, the same sync. Your phone is a first-class device, not a viewer.
Store status: the iOS and Android apps are coming to the App Store and Google Play — they are not live yet. The changelog and this page will say so plainly when that changes.
What’s the same everywhere
- Notes, canvases, memory types, tags, and properties.
- On-device semantic embedding and auto-linking — multilingual, offline.
- On-device voice transcription.
- Sync through your cloud folder or WebDAV server.
- The privacy architecture: no telemetry, gated egress, Network Audit.
Platform differences, stated plainly
| iPhone / iPad | Android | |
|---|---|---|
| AI chat | Apple Intelligence — Apple’s foundation models, system-managed* | Bring your own cloud key, or chat off — no local chat model |
| Semantic search & linking | On-device | On-device |
| Voice transcription | On-device | On-device |
| Image captions | Apple’s on-device multimodal models* | On-device |
| Cloud chat with your own key | Optional, opt-in | Optional, opt-in (it’s the only chat option) |
| License | iOS license ($11.99) or Cross-Platform | Android license ($7) or Cross-Platform |
* The Apple Intelligence dependency: on iPhone/iPad, chat and image understanding run through Apple’s foundation models under Apple’s privacy guarantees — on-device, with Apple’s Private Cloud Compute where the system chooses. That requires an Apple Intelligence-capable device (A17 Pro-class iPhone or a supported iPad). On older devices the app explains exactly why and offers cloud chat with your own key instead.
Two honest boundaries: you can’t load custom local models on any phone or tablet — on-device AI there is Apple’s models (iOS) or our bundled voice/embedding/caption models; picking your own local model is a desktop capability (via your own Ollama server). And the Android chat gap is architectural, not a pricing tier: current Android hardware plus our in-process runtime can’t run a chat-grade local model acceptably, and we won’t pretend otherwise — the Android license costs less partly for that reason. Cloud chat on any platform uses the same consent screen and endpoint disclosure as desktop.
Model downloads and Wi-Fi
Model downloads on mobile are Wi-Fi-only by default (change it in Settings if you want cellular). The embedding model (~470 MB) is required for semantic features; voice adds ~1 GB when you first use it.
Built for thumbs
The mobile UI is not a shrunken desktop: primary controls live in the bottom thumb zone on phones, edge rails on tablets, and touch targets stay comfortably large. Quick capture, search, and chat are all reachable one-handed.
Devices and slots
A mobile device occupies a mobile slot on your account (Android and iOS licenses include one each; Cross-Platform includes three). Manage and deactivate devices from your account page.