Speak, and Pour writes it down. Turn thoughts into words you can keep and send anywhere.
Pour is a free, open-source voice dictation app for Android. Fully offline. No accounts, no cloud, no subscriptions. Your words stay on your device.
- Voice dictation — tap to talk, Pour transcribes in real time using an on-device speech model
- Bucket — a quick-capture inbox for stray thoughts
- Projects — organize pieces into projects with folders
- Themes — Pour Light, Pour Dark, Sepia, Night, Forest, Ember, and custom
- Export — Plain Text, Markdown, Fountain (for screenwriters), HTML, and PDF
- Trash — deleted pieces go to Trash first, with configurable auto-delete
- Backup & restore — full library backup to a single file
| Capture | Bucket | Projects |
|---|---|---|
 |
 |
 |
| Project detail | Share | Themes |
|---|---|---|
 |
 |
 |
Pour is a native Android app (Java, no Gradle — builds with a shell script).
Prerequisites:
- JDK 17
- Android SDK (platform android-35, build-tools 35.0.1)
Build:
ABIS=arm64-v8a ./build-release-apk.shThe APK will be at pour-1.0.25-arm64-v8a-release.apk.
Signing: The release build expects a keystore. See build-release-apk.sh for details.
Pour uses the Kroko Zipformer speech recognition model, which downloads on first launch. The model is licensed CC-BY-SA. Speech recognition is powered by Kroko.
Pour is free software: you can redistribute it and/or modify it under the terms of the GNU General Public License as published by the Free Software Foundation, either version 3 of the License, or (at your option) any later version.
See LICENSE for the full text.
Pull requests welcome. Please keep the code style consistent (native Java, programmatic UI, no external dependencies beyond sherpa-onnx).