Frequently asked questions

Correct. SnoreCam has no servers, no cloud storage, and no upload code path. Your clips live encrypted on your phone and stay there. The only way a clip leaves your device is if you tap Share — and then iOS sends a copy to whatever app you choose (TikTok, Messages, etc.). SnoreCam itself still doesn't see it.

Regular snore apps record audio and surface a score and a chart. SnoreCam does both of those — Snore Score (0–100), per-minute intensity timeline, 7-night trends — plus the unique part: short video clips when audio or motion triggers fire, each one captioned in plain English by an on-device AI. You also get audio-only playback per clip if you'd rather just listen. Everything stays on the phone.

Yes — SnoreCam is a snore recording and snore detector app for iPhone, with one big difference: it records video, not just audio. When it detects snoring it saves a short video clip and captions it with an on-device AI, so you can actually see and hear what happened instead of just reading a waveform. It also computes a nightly Snore Score (0–100) and an intensity timeline. Everything stays on your phone.

Yes. SnoreCam works as a sleep talk recorder: its on-device sound recognition flags sleep talking (alongside snoring and coughing) and saves a captioned video clip of the moment. Because the camera is rolling, you get the video and the audio together — and you can play just the audio if you'd rather only listen. Nothing is uploaded.

That's exactly what SnoreCam is for. Prop your iPhone on the nightstand, tap Start, and it records short, AI-captioned video clips whenever you snore, sleep-talk, or cough through the night — then shows you a 3–5 clip highlight reel in the morning. Unlike a generic screen or video recorder, it only saves the moments that matter, and the footage never leaves your device.

A 0–100 number we compute from how long you snored and how loud it was. Roughly: light snorers land around 20, moderate around 50, heavy around 80+. It runs entirely on-device from the mic stream — same number every time for the same audio. Trends show your last 7 nights so you can see whether changing a habit (less alcohol, side sleeping, a CPAP appointment) actually moves the line.

Yes. The morning review surface includes a horizontal timeline of the whole night with intensity bars and a tappable dot for every recorded clip. Tap a dot to play that clip; the chart shows you the loud stretches and the quiet ones at a glance.

Yes. Each clip in the morning review has a headphones button that plays just the audio — no video, no warm-up — via the standard iOS audio player. Useful if you want to scrub through a snore quickly without staring at a poster frame.

Plug the phone in. Mic + motion monitoring overnight uses ~30–40% of a typical iPhone battery; with the camera only triggering occasionally, you'll wake up at 70–80% if you forget to charge. We recommend keeping it plugged in anyway.

Back camera (rear main wide). Better low-light sensor than the front, and the wide angle captures the whole bed from a nightstand 2–3 feet away. Prop your phone with the screen face-down on the nightstand and the lens aimed at your bed.

Yes — iPhone's rear camera has surprisingly good low-light sensitivity. Even in a dark bedroom with curtains closed, clips show enough detail to see body movement, position changes, and getting out of bed. A small night-light or hallway light helps the AI captioner identify scenes more reliably.

Your first 3 monitoring nights are free, with no credit card — and they don't have to be consecutive, so there's no clock running. After that it's $9.99/month or $59.99/year (about $5/month if you go annual — a 50% saving). Live preview and live captions are free forever. Cancel anytime in iOS Settings → [your name] → Subscriptions.

SnoreCam is not a medical device and does not diagnose sleep apnea or any other condition. If a partner has told you you stop breathing in your sleep, see a sleep physician — SnoreCam clips might be useful to bring along to your appointment, but the diagnosis is the doctor's, not the app's.

When you tap Share, iOS opens the native share sheet and you pick a destination. SnoreCam renders a copy of the clip with the AI caption overlaid and a small SnoreCam watermark, then hands the file to TikTok's (or Instagram's, or Messages') app. From that point on, the receiving app handles the upload — SnoreCam never sees it. You can disable the watermark in Settings if you prefer.

Only with their explicit, informed consent. Recording laws vary by jurisdiction — some require all-party consent for audio recording. SnoreCam is intended for self-monitoring; using it to surveil others without consent likely violates both our Terms and your local privacy laws.

SnoreCam's motion monitor detects significant phone movement and logs it as an event. In the morning you'll see a clip from the moment the phone moved (assuming the camera triggered) plus a note in the highlight reel that the phone was repositioned. Your session keeps running unless the phone disconnects entirely.

Yes, after 14 days by default. You can change the retention period in Settings (7, 14, 30, or 90 days). Clips you've favorited (starred in the highlight reel) are excluded from auto-deletion. Total clip storage is capped at 2 GB — the oldest unfavorited clips are deleted first if you approach the cap.

Yes, entirely. The vision-language model that writes the captions (MiniCPM-V 4.6) is bundled inside the app — there's no model download and no server. Captioning runs on-device from first launch and works in airplane mode. You can verify it yourself: turn on Airplane Mode during setup and watch the live captions keep going.

If you allow it during onboarding (or later, from Settings → Apple Health), SnoreCam writes the duration of each monitored night — bedtime and wake time only — to Apple Health. That means your sleep nights show up in the Health Sleep tile and feed sleep trend math in other apps. We never read anything from Health. No audio, no video, no captions, no Snore Score ever leaves SnoreCam for Health (or anywhere else).

Not in v1. It's on the roadmap for a later update if there's clear demand from users. The wedge for v1 is the iPhone camera, so the engineering goes there first.

Any iPhone running iOS 26.0 or later. There's no specific chip requirement — the AI model ships inside the app and runs across the supported range. Newer chips generate captions faster (a few seconds vs. several), but snore detection, video capture, and captioning work on every supported iPhone.

Yes — the audio trigger doesn't know whose snoring it is. If you sleep with a partner who snores, you'll capture both of you. If you want partner-blur on shared clips, use the face-blur toggle (coming in v1.1) before sharing.

SnoreCam captures it. Sleepwalking, sleep talking, getting up to the bathroom — these are all motion or audio triggers that save a 30-second clip from the always-watching camera. Many users find the morning highlights of these events genuinely useful to share with a sleep specialist or partner who's been telling them about it for years.

It happens. The morning review modal warns you to watch each clip before sharing for exactly this reason. The face-blur toggle (v1.1) helps but isn't a substitute for watching the clip first. Default behavior: clips stay on your phone forever — they only leave if you choose to share them.

SnoreCam uses Apple's SoundAnalysis framework with a snoring-specific classifier — the same audio-ML stack Apple ships across iOS. In our internal testing on real overnight sessions, true-positive snore detection runs ~90–95% with false positives under 5%. False positives are usually loud breathing, partner snoring, or HVAC. You can tune sensitivity in Settings if you're getting too many or too few clips.

SnoreLab is the dominant audio-only snore tracker — they're great at audio recording, snore scoring, and trend analysis. SnoreCam does all of that (Snore Score, intensity timeline, 7-night trends, audio-only playback) plus the unique part: 30-second video clips when audio or motion triggers fire, each captioned by an on-device AI. If audio alone is enough, SnoreLab is a perfectly good choice. If you want to see what's happening — position changes, getting up, mouth open — SnoreCam is the only iPhone option.

Sleep Cycle and Pillow are sleep-stage trackers with smart alarms; snoring is a secondary feature for both. SnoreCam is the opposite — snoring and video capture are the core, sleep stages come from Apple Health if you have an Apple Watch. If you mainly want a smart alarm and sleep cycle graph, use Sleep Cycle. If you want to know what your nights actually look like, use SnoreCam.

Smart alarms are a v1.x feature, not V1. We focused V1 on the differentiator — AI-captioned video clips with on-device privacy — rather than rebuilding what every sleep tracker already does. iOS 26's native Sleep Schedule already covers smart-alarm needs for most users.

Two reasons. First, it would force us to either stream audio (which conflicts with our "no servers" promise) or bundle large audio files (which bloats the app). Second, you almost certainly already have a sleep-sounds app — Calm, Headspace, Spotify sleep playlists, iOS's built-in Background Sounds. We're choosing not to duplicate something dozens of apps do well.

Yes. Each clip is a standard .mov file you can share via the iOS share sheet (AirDrop to your Mac, save to Photos, send via Messages, etc.). The caption travels with the clip as an overlay. For nightly summary data, take a screenshot of the morning review. CSV export of session metadata is on the v1.x roadmap if users ask for it.

Not in V1. The wedge is iPhone on a nightstand. iPad support would require rethinking placement and isn't on the immediate roadmap.

No. SnoreCam runs on any iPhone with iOS 26.0+, with no chip-family requirement — the AI model is bundled in the app and runs across the supported range. Newer chips simply produce captions faster; snore detection, video capture, and captioning work identically on every supported iPhone.

No. SnoreCam sends one local notification, in the morning, after your session ends: "Your night is ready." It never wakes you mid-sleep. The notification is scheduled locally on your phone (not pushed from a server) and contains no clip content.

Yes, entirely. Monitoring, audio detection, video capture, and AI captioning are all on-device. The only thing that requires network is initial App Store subscription verification, which iOS handles in the background on its own schedule.

Yes. Live preview and live captions — pointing the phone at the room and watching the on-device AI describe what it sees — are free and unlimited, forever. On top of that, your first 3 full monitoring nights are free with no credit card. SnoreCam is subscription-funded (there are no ads and no analytics — a free-with-ads model would undercut the "no servers, no tracking" promise), but you can try it, and keep using preview, without ever paying.

No. The first 3 monitoring nights are free with no card and no trial countdown — they don't even have to be consecutive, so nothing expires. You only see a subscription screen when you start a 4th night, or if you choose to subscribe from Settings. Nothing charges you automatically.

Everything. Each free night is full-feature SnoreCam — full-night monitoring, complete Snore Score history, all clips with AI captions, sharing, Apple Health write, and the storage cap up to 2 GB. After your 3 nights, starting another night asks you to subscribe ($9.99/month or $59.99/year). Anything you already captured stays yours.

Absolutely not. We can't. SnoreCam has no servers — your clips, audio, and AI captions never leave your iPhone. We have no way to collect them. The on-device vision-language model (MiniCPM-V 4.6) was trained externally by its authors and ships frozen inside the app; it never learns from your data.

iOS requires apps to request camera permission up front, even for triggered/conditional use. SnoreCam genuinely only turns the camera on when a sound or motion trigger fires; it's off the rest of the night. You can verify by watching the iOS green camera-active indicator — it only lights up during the ~30-second clip recording windows.

Android's on-device AI ecosystem isn't yet mature enough for the privacy promise that defines SnoreCam — particularly the vision-language model side. As Google's Gemini Nano matures into a comparable on-device VLM, Android becomes feasible. Realistic timeline: 2027+.