Privacy Policy

Musaic is built to keep your work on your device. We collect as little as possible, never sell data, and never use advertising or cross-app tracking. Two features need external AI services: automatic lyrics and scenes you describe in your own words. This policy names those services and explains exactly what happens to your information.

What stays on your device

Your projects, recordings, video, audio, lyrics and all edits are stored locally on your iPhone. Nothing is uploaded unless you use a feature described below, or export or share a project yourself.

Anonymous usage analytics

To understand which features are used and to improve the app, Musaic collects anonymous product-interaction analytics through TelemetryDeck (EU-hosted). This data:

• contains no personal information - no names, no emails, no lyric text, no file names, no contents of your projects;
• uses an anonymous identifier derived on-device from Apple's identifier-for-vendor and a per-app salt - it cannot be traced back to you;
• does not use the advertising identifier (IDFA), does not fingerprint you, and is not used for tracking under Apple's definition;
• can be turned off at any time in Settings → Help improve Musaic.

Lyrics (Musaic Pro)

Automatic lyrics use external AI services. From version 2.1, Musaic asks for your permission before it sends any audio for this, and our legal basis is that permission. Earlier versions send the audio when you tap Generate. You can withdraw your permission at any time under Settings → Privacy. Musaic then asks again before it sends anything. Withdrawing doesn't reach back to audio already sent. Without your permission Musaic sends no audio for lyrics, and the rest of the app keeps working.

When you generate lyrics, this happens:

1. Musaic sends the audio you picked, the whole project or one layer, over an encrypted connection to our lyrics service (api.lyrics.likwest.com), hosted at Hetzner in Finland. It can include anything sung or said near your iPhone. On most iPhones Musaic removes the music on the device first. Otherwise our service removes it. A few seconds of the unseparated audio go ahead of the rest, so you see the first words sooner.
2. Our service sends the audio to Groq, an external AI speech service, which turns it into text. If Groq can't answer, our service sends it to DeepInfra instead, and if DeepInfra can't answer, to fal. Both are external AI speech services too.
3. Our service sends the recognised words, without the audio, to Nous Research, an external AI language service, which corrects them.
4. The result returns to your device.

What each service keeps:

• Our lyrics service. We delete the audio about a minute after the job ends. The recognised text stays in the server's memory until newer jobs replace it or the server restarts. Our server logs record the IP address of each request, with no audio and no text. The access logs are deleted after 14 days. The application logs are overwritten by newer entries, so they have no fixed period.
• Groq. We have turned on Zero Data Retention for our account, so Groq keeps neither the audio nor the text. Its terms don't allow it to train models on them.
• DeepInfra, only if Groq can't answer. DeepInfra's policy says it holds the audio in memory while it works, writes it to no disk, deletes the text once it has returned it, and doesn't train models on either.
• fal, only if Groq and DeepInfra can't answer. We have told fal not to keep the audio and the text in its request history. That setting covers the request history only. fal's terms allow it to use anonymized or aggregated data derived from them to develop its products and AI models.
• Nous Research. It receives the recognised words and no audio. Its policy allows it to keep this text and use it to train models unless a customer turns on its Privacy Mode.

Audio sent before 30 October 2026 went out under these services' default settings. fal can hold it for up to 90 days, and Groq can hold it in its logs for up to 30 days.

Groq, DeepInfra and fal are based in the United States, so your audio leaves the EU. Nous Research's policy doesn't say where it processes text, so assume that happens outside the EU as well. We do not pass your IP address to any of them. We use the audio only to write the lyrics, and we do not train anything on it.

Chords and beat

Musaic works out chords and beats on your iPhone. If the models for this are not on your iPhone yet, for example while they download, Musaic sends the take over an encrypted connection to our lyrics service instead, without asking. Our service analyses it with its own models and passes it on to nobody. It deletes the audio as soon as the analysis ends. This applies to Musaic Pro, and to audio projects, which detect chords without Pro.

If you don't generate lyrics and the chord models are on your iPhone, no audio leaves your device.

Animations (Musaic Pro)

The animated scenes are drawn on your device. To find the objects for a scene, the text of the sung line is sent over an encrypted connection to our catalogue service (api.lyrics.likwest.com), which matches it against its own catalogue on our server and answers with the drawing instructions. No audio and no video are sent for this.

If you describe a scene in your own words instead, that description is sent to the same service, which forwards it to two external AI language services, fal and Nous Research. Musaic asks before the first time. Write nothing there you would not hand to a third party. We do not use the lines or your descriptions for analytics, and we do not train anything on them.

Feedback

After a few exports Musaic may ask whether you enjoy it. If you answer "Not really", you can write a short message. It is sent over an encrypted connection to our server (api.lyrics.likwest.com) together with the app version, the iOS version and your language and region setting. It carries no name, no email address and no device identifier, and we do not store your IP address with it. We read these messages only to improve the app. They reach the developer through a private Telegram chat and are not shared with anyone else. Please don't write anything personal into the message. If you want a reply, use the email address under Contact instead.

Links you save as Inspo

When you share a link into Musaic or add it as Inspo, Musaic shows its title and a preview image. To get them, your iPhone asks the platform the link belongs to: Apple Music (itunes.apple.com), Spotify, Deezer, TikTok, Instagram or YouTube. For any other link it loads the linked page itself. The request goes straight from your device to that platform, not through our servers. It contains the link and, like every web request, your IP address. It carries no login and nothing from your projects. The platform's own privacy policy covers what it does with the request. If you don't want this, don't save links as Inspo.

Subscriptions & purchases

Musaic Pro is sold through Apple's App Store using StoreKit. Apple handles the payment and your billing information - we never see or store your payment details. The app only learns whether an active subscription exists, in order to unlock Pro features.

Camera, microphone & photo library

Musaic uses your camera and microphone to record, and your photo library to import and save videos. These are used only when you actively record, import or export. Nothing is captured in the background, and nothing is sent to us except through the features described above.

Children

Musaic is not directed at children under 13 and does not knowingly collect personal data from them.

Your rights

Because we don't hold personal data linked to your identity, there is no account to delete. You can disable analytics and withdraw your permission for AI services in Settings, and deleting the app removes all local data. You also have the right to access, correct and delete your data and to complain to a data protection authority. For any questions, contact us below.

Changes

If this policy changes, we'll update the date above and post the new version on this page.

Contact

musaicapp@outlook.com