Help & support.
A multitrack recorder that reports nothing back to Silo. Here's how to get help with it.
Email a human.
direct · no ticket queueEmail us about a bug, a multi-mic setup question, cuts, punch-in, the publish folder, or a feature request. Your episodes live on your Mac, so please don't attach episode audio, transcripts or voice files to your message; we can help without them. For non-private questions you can also open an issue on GitHub.
Email support ↵The short answers.
privacy · recording · cuts · voicesNo. Recording, transcription, cuts, renders, search and publishing run on your Mac, with no account and no telemetry. Episode packs (titles, show notes, social posts) are drafted on your Mac too, unless you opt in to your own cloud AI key in Settings. Then only the episode title and the transcript text go to the provider you chose. Audio is never sent.
Yes. Select several Core Audio input devices or interfaces and Silo Cast records one 44.1 kHz mono track per selected device. A multi-channel interface is mixed down into its one device track rather than split into individual voices. If a device drops out mid-take, the audio it already captured is kept as a track marked partial.
Silo Cast asks which engine to use. Apple's speech engine, with on-device recognition required, needs no download. Or approve a one-time download of an optional on-device speech model, about 1 GB, from a pinned and checksum-verified source; it also runs on your Mac and can be removed in Settings. Either way, your audio is never sent anywhere.
No. Correcting a speaker label or wording saves a new text revision; Transcript edits do not alter, cut, or replace the recorded source audio. Cutting is its own deliberate step: tick transcript lines, review each cut, preview the join, then render a new M4A beside a JSON edit list. The original recordings are never rewritten.
Yes, with a voice you made in Silo Clone (coming soon, macOS 26). Silo Cast does not record or create voices. Export a .silovoice file from Silo Clone, open it in Settings, and approve the 408 MB voice model download. Then select the transcript lines, type the corrected English wording and audition the take before you accept or discard it. A voice file saved without Silo Clone's rights confirmation is refused. Replacement takes carry no watermark, so disclose them where that matters.
On Macs with Apple Intelligence, titles, show notes and social posts are drafted on-device. On other Macs you get an excerpt-based draft built from the real transcript and timestamps. Chapters are always built on your Mac. If you opt in to your own cloud key in Settings, the provider you chose writes the draft instead, and Silo Cast labels it as cloud-written.
One publish action writes a WAV or M4A mixdown, chapters, SRT and VTT captions of the reviewed transcript, show notes, a release checklist and a SHA-256 manifest into a new folder you choose. The folder appears only once every file is written and never overwrites an earlier one.
Its clips are audiograms: up to three minutes of the episode shown as its waveform, your episode artwork and the speaker's name, with the reviewed transcript lines on screen. For captions on your own video footage, caption styling or caption translation, use Silo Caption (coming soon, macOS 15).
Silo Cast is a one-time purchase of $19.99 on the Mac App Store. No subscription, no per-episode fee and no upload quota: buy it once and record as much as you like.