JJevis
PRIVACY POLICY

Privacy, in plain language.

This policy explains what Jevis processes, why it is needed, and the choices available to you.

Last updated September 23, 2026
At a glance
01

You control capture

Jevis receives screen frames only while you run the iOS screen broadcast.

02

OCR stays on device

Apple Vision recognizes text locally. Jevis does not upload raw screen images to its own server.

03

Chat text powers AI

Recognized conversation context is sent to configured AI endpoints to produce analysis and reply suggestions.

1. Scope

This policy applies to the Jevis iOS app, its keyboard extension, and its screen-broadcast extension. It does not replace the privacy policies of Apple, RevenueCat, DeepSeek, TypeSafe AI, or another model endpoint you choose.

2. Information processed

Screen content: while you actively run the JevChat broadcast, recent screen frames are copied into the app's shared container so the app and keyboard can read the visible conversation.

Recognized conversation data: on-device OCR extracts visible text. Jevis may derive speaker position, recent message context, intended recipient, and related reply instructions.

Model configuration: model names, endpoint URLs, and API credentials that you enter are stored locally in the app's shared container. Credentials are included in requests to the endpoint you configure.

Purchase information: RevenueCat processes information needed to check products, transactions, and Pro entitlement status. Apple processes App Store purchases.

Technical data: iOS and connected services may process basic request, device, network, and diagnostic information under their own policies.

3. How information is used

Jevis uses this information to recognize visible chat messages, analyze conversation intent, generate reply suggestions, maintain app settings, and verify paid access. We do not intentionally sell personal information, build advertising profiles, or use chat content for advertising.

4. Where processing occurs

Raw screen frames and OCR run on your device. Recognized chat context and prompts are transmitted over HTTPS to the configured AI services—by default, TypeSafe AI for conversation analysis and DeepSeek for reply generation—or to another endpoint you select. Those providers act under their own terms and may process or retain requests in other countries.

5. Storage and retention

Jevis does not operate a user-account database or a first-party backend for storing your chats. Recent frames, settings, credentials, reply state, and related app data may remain in the iOS app group until replaced, cleared by the app or operating system, or removed when you delete the app and its extensions. Third-party providers set their own retention periods; review their policies before sending sensitive content.

6. Your choices

You can stop the screen broadcast at any time, avoid opening sensitive conversations while broadcasting, change or remove model credentials, choose another compatible model endpoint, and delete Jevis to remove locally stored app data subject to normal iOS behavior. Purchase records remain with Apple and RevenueCat as required for transactions and entitlements.

7. Security

Jevis uses iOS app-group protections and HTTPS requests to supported endpoints. No system is perfectly secure. Protect your device and API credentials, and do not submit secrets or highly sensitive conversations unless you accept the configured provider's practices.

8. Children

Jevis is not designed for children below the minimum age required to consent to digital services in their location. If you believe a child has provided personal information through Jevis, use the support channel below.

9. Changes

We may update this policy as Jevis changes. The date at the top shows the latest revision. Material changes will be reflected on this page.

10. Contact

For privacy questions or requests, use the Jevis support page. Do not include API keys or private chat content in a support request.

Open Jevis Support