OLI Privacy Policy

Effective date: September 21, 2026

OLI ("the app," "we," "our") is a push-to-talk voice translator. This policy explains what happens to your data when you use it.

The short version

OLI recognizes speech, translates it, and speaks the translation back to you entirely on your device. We do not collect, transmit, sell, or share your voice, your translations, your usage history, or any other personal data. The app has no account, login, or sign-up of any kind. The only outbound network requests OLI makes are plain file downloads for optional language packs or the optional Whisper recognizer (see below) and, if you buy Offline Pro, the purchase itself, handled entirely by Apple. None of those requests carry your voice, your translations, or any content you've spoken.

What data OLI accesses on your device

• Microphone audio. When you hold the push-to-talk button, OLI records audio to recognize and translate what you say. That audio is kept in memory only for the length of the current turn (capped at 30 seconds) and is discarded as soon as the turn finishes. It is never written to disk, never saved as a transcript, and never sent to us or anyone else.
• Translated text. The text OLI recognizes and translates is processed on-device and shown on screen for that turn. OLI does not keep a history or transcript of past conversations.
• Rating prompt timing. To decide when it's reasonable to ask you for an App Store rating, OLI stores a small amount of technical metadata on your device: how many turns you've completed and when the app was first opened and last prompted. This is aggregate counting and timing information only, never your voice or translations, stays on your device, and is never transmitted anywhere.

Language pack and Whisper downloads

OLI ships with a base set of languages built in. Additional language packs, and the optional multilingual Whisper Base Q5 speech recognizer, can be downloaded from within the app. Each of these is a plain file download over HTTPS, verified against a known SHA-256 checksum before it's used, the same as downloading any file from a website. The request contains no personal data: not your voice, not your translations, not any account or advertising identifier, because OLI has none to send. Once downloaded, a language pack or the Whisper recognizer runs entirely on your device like everything else in the app.

What OLI does not do

• We do not require or request an account, email address, or sign-in of any kind.
• We do not use third-party analytics, advertising, or tracking SDKs.
• We do not transmit your voice, translations, or usage data to us or to any third party. OLI has no backend server of its own to send it to.
• We do not keep a transcript history, on your device or anywhere else.
• We do not build a profile of you, and we cannot, because we never receive your data in the first place.

Offline Pro purchase

Offline Pro unlocks downloadable language packs and the optional Whisper recognizer. It's available two ways: Offline Pro Yearly, an auto-renewing subscription billed once a year until cancelled, and Offline Pro Lifetime, a one-time, non-consumable purchase with no recurring charge. Purchases, renewals, and restores are handled entirely by Apple's App Store through StoreKit. We do not receive or store your payment information, and OLI does not use an app account or a backend of its own for purchases. Apple's own privacy policy governs that transaction.

If you choose Offline Pro Yearly, payment is charged to your Apple ID at confirmation of purchase and renews automatically each year unless auto-renewal is turned off at least 24 hours before the end of the current period. You can manage or cancel the subscription any time in Settings → [Your Name] → Subscriptions on your iPhone. See Apple's Terms of Use for the full subscription terms.

Children's privacy

OLI does not knowingly collect personal data from anyone, including children, because it does not collect personal data from any user. The app is a general-purpose translation tool and is not directed at children specifically.

Third-party and open-source components

OLI uses on-device speech recognition, translation, and speech synthesis components, including MIT-licensed software such as whisper.cpp / whisper.rn and the converted Whisper model from the ggerganov/whisper.cpp project. These run entirely on your device once installed and do not transmit any data anywhere on their own.

Changes to this policy

If OLI's data practices change, for example if a future version adds an account system, cloud sync, or analytics, we will update this policy before that version is released and update the effective date above.

Contact

Questions about this policy: [email protected]