Help & support.
M4B audiobooks from your own files, the sentence-by-sentence reader, voices, and your local library.
Email a human.
direct · no ticket queueTell us which device and file type you used, what you expected, and what happened. Please do not attach a sensitive source document; a redacted sample or short description is usually enough.
Email support ↵The short answers.
privacy · formats · voices · exportsImport, OCR, search, narration, transforms, citations, and audiobook assembly run on your device, and Silo Systems receives none of that content. On Mac, optional voice and model downloads go directly to the pinned source named in the privacy policy, only after you approve their size; that host sees ordinary request details such as your IP address, never your documents or questions.
PDF, EPUB, DOCX, PPTX, RTF, saved web pages, plain text, text-bearing DAISY books, and photos or scans of printed pages, including files picked from any Files provider. Scanned pages are recognized on the device after each page's language is detected, and every sentence you import becomes searchable.
MP3, M4A, M4B, and other system-supported audio. Audio-only or synchronized DAISY packages are not supported.
Yes. Imported original audio keeps an exact listening position across launches. It can continue in the background, respond to lock-screen and headset controls on iPhone and iPad, and resume from the library without changing the source file.
Apple system voices start reading immediately with no download. The recommended neural reading voice is included on iPhone and iPad; on Mac it is a one-time 336 MB download that starts only after you approve its size and license. Neural voices speak English; other languages read with Apple system voices.
On Mac, yes, if you made the voice in Silo Clone (coming soon, macOS 26). Silo Read does not record audio, ask for the microphone, or create voices. Export the voice from Silo Clone as a .silovoice file, then in Silo Read open the voice picker and choose My Voices, Import from Silo Clone. A file saved without Silo Clone's rights confirmation is refused. Speaking in an imported voice needs the optional personal-voice engine, about 3.2 GB, installed from Settings after you approve it; synthesis then runs entirely on your Mac. Personal voices are not available on iPhone or iPad.
A single M4B assembled with cover art, audiobook metadata, and native timed chapter markers; a whole document or one chapter as AAC M4A; and document text or existing chapter recaps as Markdown or plain text. It does not claim MP3 export or fabricate word-level read-along timing.
Silo Read retrieves passages from your local library and validates document, chapter, page, and sentence references after an answer is written. On iPhone and iPad, answers use the system's built-in on-device model where the device supports it; on Mac, they use an optional local text model download. If no model is available or the evidence is insufficient, the app shows the evidence instead of inventing an answer.
Pricing: $19.99 one-time · no subscription. One universal purchase covers Mac, iPhone, and iPad. It requires macOS 14 or later on Apple Silicon, or iOS/iPadOS 17 or later. The optional Mac text model needs at least 12 GB of unified memory, with 16 GB recommended; the optional personal-voice engine needs at least 8 GB, with 16 GB recommended.