AI Translation: No Internet (displayed on your device as AI Translation, and referred to in this policy as "we," "our," "us," or the "App") is a native mobile application for Apple iOS and Google Android designed for live, on-device speech-to-speech translation. We built the App around a single principle: your conversations stay on your device.
This Privacy Policy explains how information is handled when you use the App and why specific device permissions are requested on iOS and Android.
100% Local Processing
Speech detection, the STT AI Model, the Translation AI Model, and speech synthesis run entirely on your phone.
No Audio Collection
Your microphone audio, spoken words, and translated text never leave your device or touch any external server.
Cleared When Done
Transcripts live only in temporary memory (RAM) during an active session and are erased as soon as the session ends.
No Accounts or Analytics
No user accounts, no third-party analytics, no crash-reporting telemetry, and no profiling.
1. Summary of Key Privacy Practices
- 100% On-Device Speech & Translation: All voice activity detection, speech recognition (
STT AI Model), language translation (Translation AI Model), and speech synthesis (Text-to-Speech) run locally on your device. - No Audio or Transcript Collection: Your microphone audio, spoken words, and translated text never leave your device and are never transmitted to us or any third-party cloud server.
- Ephemeral Conversations: Session transcripts exist only in temporary device memory (RAM) during an active translation session and are permanently erased as soon as you end the session or close the App.
- No Accounts, Analytics: The App does not require user registration and contains no third-party analytics, crash-reporting, or tracking SDKs.
2. Information Processed Exclusively on Your Device
a. Microphone Audio
When you start a translation session, the App captures live audio from your device's built-in microphone (or connected headset microphone, if selected) to detect speech and transcribe spoken words on your device.
- Audio streams are processed in real time in volatile memory (RAM) and immediately discarded.
- The App never records, saves to disk, or transmits your audio externally.
b. Conversation Transcripts and Local Context
During an active session, recognized speech and translated text are displayed on your screen. To improve translation accuracy across sentences, the App may temporarily pass a small number of recent turns (configurable between 0 and 6 turns in Settings) to the on-device Translation AI Model as read-only local context.
- When you tap End to finish a session or close the App, all transcripts and conversation context are immediately cleared from device memory.
- No conversation history or transcript log is saved on your device or elsewhere.
c. Local App Settings and Downloaded AI Models
The App stores your configuration preferences locally on your device (using Android SharedPreferences or iOS UserDefaults), including:
- Your selected primary language and partner language
- Session mode (
Listening,Conversation, orCustom) and audio output routing preferences (Headphones,Device Speaker, orMuted / Text Only) - Microphone input preference, automatic playback speedup setting, and context turn count
- An optional user-supplied authentication token if you configure one to download AI models
- Downloaded on-device AI model files stored in the App's private internal storage
You can delete downloaded AI models at any time from the App's Settings screen or erase all local preferences and models by uninstalling the App. Depending on your device settings, your operating system (Apple iCloud Backup or Google Backup) may back up standard app preferences under Apple's or Google's respective privacy terms.
3. Network Access & AI Model Downloads
The App requires an internet connection only to download the on-device STT AI Model and Translation AI Model files during initial setup (or if you re-download them later in Settings). Once the models and offline system voices are installed, the App operates 100% offline, including in Airplane Mode.
When your device downloads AI model files:
- Files are fetched directly over HTTPS from third-party model hosting infrastructure.
- As with any standard internet download, the hosting provider automatically receives standard HTTP request metadata (such as your IP address, device user-agent, file request URL, and your optional Hugging Face access token if you manually entered one).
System offline Text-to-Speech (TTS) voices are managed directly by your device's operating system (Apple iOS or Google Android) and are subject to Apple's or Google's privacy policies.
4. Device Permissions Explained
To function properly, the App requests the following system permissions:
- Microphone (
RECORD_AUDIOon Android / Microphone Access on iOS): Required to capture spoken audio for real-time on-device speech recognition and translation. Audio is processed only while a translation session is actively running and not paused. - Speech Recognition & TTS Services (iOS Speech Recognition / Android TTS & Speech Service Queries): Used to support on-device speech processing and to check whether offline system Text-to-Speech voices are installed on your device.
- Bluetooth & Audio Routing (
BLUETOOTH,BLUETOOTH_CONNECT,MODIFY_AUDIO_SETTINGSon Android /AVAudioSessionon iOS): Used solely to detect whether wired or Bluetooth headphones are connected and to route translated audio between your headphones and the phone speaker according to your selected session mode. We never use Bluetooth permissions to determine your physical location or scan for nearby unrelated devices. - Foreground Service, Wake Lock & Notifications (
FOREGROUND_SERVICE,FOREGROUND_SERVICE_DATA_SYNC,WAKE_LOCK,POST_NOTIFICATIONSon Android): Used exclusively while downloading large on-device AI model files so downloads can complete reliably if you switch apps or lock your screen, and to show download progress notifications.
You may revoke any permission at any time in your device's system Settings, though revoking Microphone access will prevent the App from capturing speech for translation.
5. Data Sharing and Third-Party Disclosures
Because we do not collect, transmit, or store any personal data on external servers:
- We do not sell, rent, trade, or share your personal information or audio data with any third parties.
- We do not use your conversations or audio to train AI models.
6. Children's Privacy
The App does not knowingly collect personal information from anyone, including children under the age of 13 (or the applicable age of digital consent in your jurisdiction). Because all processing occurs locally on the device without data collection, no personal data from children is stored or transmitted to us.
7. International Privacy Rights (GDPR, CCPA/CPRA, and Global Laws)
Under data protection laws such as the European Union General Data Protection Regulation (GDPR), the UK GDPR, and the California Consumer Privacy Act (CCPA/CPRA), users have rights to access, correct, delete, or restrict processing of their personal data. Because all settings and AI models reside solely on your device and we do not collect or hold any personal data on our servers, you have full, direct control over your data at all times:
- End a session to immediately erase active transcripts from memory.
- Delete AI models in the App's Settings screen.
- Clear app storage or uninstall the App via your device settings to permanently remove all locally stored preferences and files.
8. Changes to This Privacy Policy
We may update this Privacy Policy from time to time to reflect changes in the App's features or legal requirements. Any updates will be posted at this URL with an updated "Last Updated" date.
9. Contact Us
If you have any questions or concerns about this Privacy Policy or the App's privacy practices, please contact us at:
Developer / Organization: timdinh.com
Email: [email protected]