← vibephone

Privacy Policy

Last updated: 2026-09-19

This policy explains what data Vibephone (the "App") collects, why, and how it's handled. Vibephone is developed by an individual developer operating as Dawa King ("we", "us"). If you have questions, reach us via GitHub Issues.

The short version

Your voice never leaves your iPhone. Speech is recorded and converted to text entirely on your device, and the resulting text is sent straight to your own computer over a direct Bluetooth link. We operate no transcription server and never receive your audio or your transcripts.

What data we collect

We collect only what's necessary to run the app. Specifically:

1. Audio recordings — processed on-device, never uploaded. Short clips (up to 60 seconds each) that you record by tapping the mic button are transcribed locally on your iPhone using an on-device speech-recognition model (WhisperKit). The audio is written to a temporary file, transcribed, and deleted. It is never transmitted to us or to any third party.

1. Transcripts — never sent to us. The text is delivered directly from your iPhone to your paired computer over a local Bluetooth Low Energy connection. It does not pass through our servers, and we do not store, log, or train on it.

1. Pairing code — a 6-digit code shown by the Vibephone Bridge app on your computer, which you enter once on the phone. It is stored locally on both devices and is used only to stop an unwanted nearby phone from typing on your computer.

1. Subscription records — if you subscribe, we use RevenueCat to manage your entitlement. We store your tier, subscription status, and expiration date, linked to an anonymous, opaque identifier. Apple-issued purchase receipts are processed by RevenueCat on our behalf. This is the only category of data that leaves your device to us or our processors.

We do not collect: your audio, your transcripts, name, email, phone, photo, contacts, location, health data, advertising identifiers, or anything we haven't explicitly listed above.

How we use it

• Provide the service. Validate subscriptions.
• Comply with legal obligations. Respond to lawful requests and enforce our Terms.

Transcription itself requires no data collection at all — it happens entirely on your device.

We do not sell your data. We do not use your data for advertising. We do not track you across apps or websites.

Who we share it with

Because transcription is on-device, there are no speech or transcript processors. The third parties involved are:

• Hugging Face — hosts the speech-recognition model. The App downloads it once, the first time you open it (and again only if you choose the larger model). Hugging Face receives an ordinary web request for the file, which includes your IP address, as any website would. No audio, text, or identifier of ours is sent.
• Apple — processes subscription payments.
• RevenueCat — manages subscription lifecycle on our behalf (purchase verification, renewal, refund, grace-period tracking). Receives an anonymous identifier and the subscription metadata Apple provides.

Apart from these, we do not share your data with anyone.

Retention

• Audio clips: never leave your device; the temporary file is deleted right after on-device transcription.
• Transcripts: never stored by us; delivered directly to your paired computer over Bluetooth.
• Subscription records: retained as long as the subscription is active, plus up to 7 years afterwards for Apple-compliance / tax-compliance reasons.

Your rights

Depending on your jurisdiction (GDPR, CCPA, etc.), you may have the right to access, correct, delete, or export your data. Because your speech and transcripts never reach us, and any subscription record is tied only to an opaque identifier, exercising most of these rights is as simple as deleting the app. To request deletion of subscription records, open a GitHub issue and we will walk through the purge with you.

Children

Vibephone is not directed to children under 13 and we do not knowingly collect data from them. If you believe a child has provided data to us, contact us via the link below and we will delete it.

Security

Speech recognition runs locally, so your audio is never exposed to the network. Text is sent to your computer over a direct Bluetooth link between your own devices — it does not traverse the internet.

That Bluetooth link is encrypted: both devices must complete Bluetooth pairing (you will be asked to confirm it the first time), and Vibephone declines to exchange any data over an unencrypted connection. The 6-digit pairing code is a separate, additional check that decides which phone is allowed to type on your computer.

Subscription traffic to our processors is encrypted via TLS. We do not process payment card data — that's handled entirely by Apple.

Changes

We may update this policy; the "Last updated" date at the top will reflect the most recent revision. Material changes will be surfaced in-app via a notice before they take effect.

Contact

Dawa King — via GitHub Issues.