Networked SDR receiver — Privacy Policy (iOS and Android)
Last updated: September 14, 2026
Demod has no accounts, no analytics, no advertising, and no developer-operated servers. Radio audio, recordings, captions, translations and decoding are processed on your device.
Demod ships as two separate apps, one for iOS and one for Android. They share the same principles, but they use different on-device components, so each has its own section below. Read the section for the app you use.
Demod for iOS (iPhone & iPad)
The iOS app does not collect, store, transmit, or share any personal data.
No analytics, no tracking, no advertising. Demod for iOS contains no third-party analytics, advertising, or tracking SDKs.
No account, no sign-in. Demod does not ask for or store any identity.
Your data stays on your device. Saved frequencies, connection details, and app settings are stored only locally on your device. They are never uploaded.
Direct connections only. Demod connects only to the SDR servers you explicitly configure (rtl_tcp, SpyServer, KiwiSDR, OpenWebRX). Those connections go directly from your device to the server you choose, so that server's operator sees your IP address and tuning requests. The developer operates no servers and receives no data about you or your usage.
Local audio, recording & transcription. Demodulated audio is played on your device (or an AirPlay device you select). If you tap Record, the audio you hear is saved to a file on your device that only you control — share or delete it yourself. If you turn on Live Text, transcription runs entirely on your device using Apple’s on-device speech recognition. Recordings and transcripts are never uploaded or transmitted by Demod, and the developer never receives them. Deleting the app deletes all recordings and transcripts.
On-device translation. If you turn on Translate, captions are translated with Apple’s on-device Translation framework. Language packs are downloaded from Apple once, but your audio and text never leave the device.
On-device signal identification. The Identify feature (including “ask Demod” via Siri) measures the tuned signal on your device and looks it up in a built-in band plan. On devices with Apple Intelligence, Apple’s on-device model may phrase the explanation. Nothing about what you tune or listen to leaves your device.
On-device decoders. The CW, RTTY, PSK31, and FT8 decoders run entirely on your device, turning the audio you’re already receiving into text locally. They transmit nothing, and the developer never receives anything.
Demod for Android
The Android app does not require an account, and the developer does not operate any analytics or advertising service in it. It differs from the iOS app in two ways: speech recognition uses downloadable Vosk models, and translation uses Google ML Kit, whose SDK contacts Google as described below.
Local processing. Radio audio, recording, speech recognition, translation, signal identification and digital decoding run on your device. Demod does not upload your recordings or caption text. Files leave the app only when you choose to share or export them.
Radio connections. Demod connects directly to the KiwiSDR, OpenWebRX, SpyServer or rtl_tcp receivers you select. Their operators receive your IP address and your connection and tuning requests. Some radio protocols use unencrypted connections.
Directories and maps. To show public receivers and map tiles, Demod makes requests to Airspy (SpyServer directory), the KiwiSDR community directory and its CDN, ReceiverBook (OpenWebRX directory) and OpenStreetMap. Reachability checks contact the listed receivers directly. These services receive your IP address; their own policies apply. Receiver coordinates describe the radio antenna, not your phone’s location. Demod does not request device location permission.
Speech models (Live Text). Enabling Live Text downloads the selected Vosk speech model from its host (alphacephei.com). Recognition then runs locally. The model host receives your IP address and the download request.
Translation (Google ML Kit). Google ML Kit translates caption text on your device; it does not send your captions or the translated text to Google. However, the ML Kit SDK contacts Google to download translation models and for updates and diagnostics. Google documents that this includes device and app information, identifiers, performance and API usage metrics, and the languages you configure. Google’s ML Kit terms and Google’s privacy policy apply to that SDK data. Demod for Android does not use the Android advertising ID.
Storage and deletion. Settings, saved servers, memories, recordings, transcripts and downloaded speech models are stored in the app’s private storage. You can delete recordings and transcripts in the app, or clear the app’s storage in Android settings. Android may back up or transfer your settings under your device’s backup settings; recordings, transcripts and downloaded speech models are excluded from Demod’s backup rules.
Permissions. Demod for Android uses Internet (radio streams, directories, model downloads), a foreground media-playback service and notifications (to keep audio playing in the background with lock-screen controls), and wake lock (to keep streams alive). It does not request microphone, camera, contacts or location access.
Your data & rights
The developer holds no data about you on either platform, so there is nothing for us to access, correct, or delete. Deleting the app removes all locally stored settings, recordings, transcripts and downloaded models. Data held by third parties named above (receiver operators, directory hosts, Google for ML Kit) is governed by their policies.
Changes
This policy may be updated to reflect new features; material changes will be noted by the “Last updated” date above.