Last updated: November 4, 2026
Poducasto is built with the help of the following open source and third-party libraries and models. We're grateful to their authors and maintainers. Each one is used under the terms of its own license, linked below.
| Library / model | Purpose in Poducasto |
|---|---|
| whisper.cpp and ggml (Georgi Gerganov and contributors) | Runs speech recognition on your device. Bundled as native code. |
| OpenAI Whisper model weights (ggml conversion) | The speech recognition model itself, downloaded on first use. |
whisper.cpp license · Whisper license
The rest of Poducasto's foundation is licensed under the Apache License, Version 2.0.
| Library | Publisher |
|---|---|
| Kotlin standard library, Kotlin Coroutines, kotlinx.serialization | JetBrains |
| AndroidX Core, AppCompat, Lifecycle, Activity, Navigation | Google / The Android Open Source Project |
| AndroidX Compose (UI, Material 3, Material Icons Extended) | Google / The Android Open Source Project |
| AndroidX Room, DataStore | Google / The Android Open Source Project |
| AndroidX Media3 (ExoPlayer, UI, Common) | Google / The Android Open Source Project |
| Material Components for Android | |
| sherpa-onnx (on-device AI voices) | k2-fsa contributors |
| Apache Commons Compress | The Apache Software Foundation |
| Kokoro-82M voice model | hexgrad |
| Qwen3-TTS (used to create the voices of the podcast packs) | Alibaba Cloud / Qwen team |
Full text of the Apache License 2.0
Voice models from the Piper project (MIT), trained on these datasets:
| Voice | Dataset and license |
|---|---|
| Siwis (French) | SIWIS French Speech Synthesis Database, University of Edinburgh — CC BY 4.0 |
| Sharvard M / F (Spanish) | Sharvard Corpus, University of Edinburgh — CC BY 3.0 |
| Alba (English) | University of Edinburgh dataset — CC BY 4.0 |
| Ramona (German), Riccardo (Italian) | The M-AILABS Speech Dataset |
| Davefx, Faber, Tugão, Thorsten, Kerstin, Gilles, Paola | CC0 (public domain) datasets |
The voices are used unmodified, only converted to the sherpa-onnx format.
| Library | Purpose in Poducasto |
|---|---|
| ML Kit Translate | On-device subtitle translation, only when you use it. Downloads a language model via Google Play Services; the text is translated locally on your device and is never sent anywhere. |
| Google Play Billing Library | Buying the podcast packs. |
| Play Asset Delivery | Downloading the podcast packs you own. |
Distributed by Google under its own SDK terms rather than an OSI-approved open source license.
Google Play Services open source notices · Google APIs Terms of Service
| Service | Used for |
|---|---|
| Hugging Face | Hosting the Whisper model file that the app downloads on first use. See the Privacy Policy. |
| GitHub (sherpa-onnx releases) | Hosting the AI voice models you can download. |