Silo Fable
iOS app by Omar Quddusi. Entertainment · Omar Quddusi
- Store rating
- 0 / 5
- Store rating count
- 0
- Download price
- Free to download
- In-app purchases
- Unknown
- Version
- 1.1.0
- Listing last refreshed
- 2026-09-07
View the original store listing
Store description excerpt
Meet one original companion on a private voice-only animated stage. Choose Auri, Kael, Vexa, Mara, Rook, or Sable, give that selected character your own name, then hold to speak and release to send. A verified reply from Apple's built-in on-device language model can be spoken through the companion's bundled persona voice, one personal voice you create, or an installed Apple system voice. There is no keyboard composer and no cloud or canned-response substitute. SIX DISTINCT CHARACTERS * Each fictional adult character has a stable history, speaking style, visual identity, boundaries, and perspective on The Common Tide. * The Common Tide lives in character knowledge. * Change characters or edit cast names later without surrendering control of your local memory. CREATE ONE PERSONAL VOICE * In Settings, record or import one 5-15 second English reference clip. Imports are checked before decoding; longer clips require confirmation before the first 15 seconds are kept. * Creation begins with a misuse warning, a required no-default choice that the voice is yours or you have the speaker's permission, the complete versioned consent statement, and a required acknowledgement. * The optional PocketTTS CoreML model is a separate 407,972,690-byte consent-gated download. Its immutable revision, per-file byte counts, SHA-256 digests, license, and attribution are shown and verified before use. * The reference recording, profile, and derived conditioning cache remain inside Silo Fable until you delete the profile. They are never uploaded or used to train a model. * Generated personal-voice audio has no audible or inaudible watermark. Treat that as a limitation, not a safeguard. If the model or profile becomes unavailable, replies use the selected character's bundled persona voice with a visible notice. PRIVATE BY CONSTRUCTION * Choose Apple's built-in on-device Speech recognizer or the optional Qwen3-ASR local MLX model. The Qwen model is a separate consent-gated 1.02 GB download, verified file by file, and removable in Settings. * Held dictation audio remains ephemeral and is never uploaded or written to a file. Only the personal-voice reference clip is intentionally saved, after the creation consent flow, until you delete it. * Transcripts, conversations, character names, personal-voice data, and the manual memory ledger remain in the app's private local container. * No account, advertising, analytics, tracking, Silo cloud transcript, or external AI inference. * Inspect, edit, pin, forget, or reset memory whenever you choose. COMPATIBILITY * The app shell and personal-voice profile creation support iOS and iPadOS 18 or later. * Generated companion replies require iOS or iPadOS 26 or later, an Apple Intelligence-capable device, Apple Intelligence enabled and ready, downloaded Apple model assets, and a locale supported by Apple's on-device model. * Local MLX recognition requires at least 8 GB memory. PocketTTS personal-voice speech requires at least 6 GB memory,
Review coverage
AppRill has captured 0 reviews for this app. This is a collected sample, separate from the store rating count. Latest review capture: Not captured yet.
Recent captured chart positions
No chart positions have been captured for this app yet.
Chart position, rating count and review coverage do not establish downloads, revenue or profit. Understand the differences.