Updated: July 22, 2026
This notice explains the third-party open-source components, system capabilities, model files, and related license information used by SpleeterAudioStems (the "App") to provide audio/video processing, stem separation, playback, export, recording, sharing, and related features.
The App uses third-party open-source software, Apple system frameworks, audio/video processing frameworks, machine learning models, or model files converted into other formats. The intellectual property rights in these components and models belong to their respective rights holders and are subject to their applicable open-source licenses, model licenses, terms of use, or platform rules.
The App's use of third-party components does not mean that the relevant third-party rights holders recognize, sponsor, endorse, or participate in the development, operation, or promotion of the App.
The App's stem separation capability is based on or references Deezer's open-source project Spleeter and uses or supports Core ML model files converted from Spleeter pretrained models.
Project information:
Project name: Spleeter
Rights holders: Deezer SA and its contributors
Project URL: https://github.com/deezer/spleeter
License: MIT License
The Core ML stem separation model names currently used or supported by the project include:
Spleeter2stemsWaveform;
Spleeter4stemsWaveform;
Spleeter5stemsWaveform.
If model weights come from a third-party conversion repository, an individually published version, or a non-Deezer official source, the actual source and license of that model file shall apply. The scope of use of model weights should not be determined solely based on the MIT License of the Spleeter code repository.
Model conversion, format compression, or repackaging does not change the license obligations of the original model. If the original model requires copyright notices, license text, source attribution, or other attribution marks to be retained, the App will continue to retain them.
Spleeter and the App provide audio analysis, separation, and processing tools. They do not automatically grant users any rights to use third-party music, recordings, videos, accompaniment, vocals, drums, bass, piano, or other track content.
Before importing, separating, recording, exporting, composing, or sharing audio/video content, users should ensure that they have lawful rights to the relevant content or have obtained sufficient authorization from the rights holders.
Separated tracks may still contain protected expression from original works, sound recordings, performances, or video content. Separating tracks into vocals, accompaniment, or other stems does not mean that the content may be freely used commercially, distributed, adapted, or republished.
The App uses FFmpeg, FFmpegKit, or related FFmpeg dynamic libraries to provide audio/video analysis, transcoding, audio extraction, audio/video composition, and related features.
The current project integrates FFmpegKit / FFmpeg as an LGPL build and retains the corresponding LGPL license text and related open-source component license texts in the project. FFmpeg, FFmpegKit, and bundled external libraries are owned by their respective rights holders and are subject to their applicable open-source licenses.
Based on the current project implementation, the App also uses the following third-party components or system capabilities.
Name
Purpose
License / Source
SnapKit
iOS UI Auto Layout
MIT License / https://github.com/SnapKit/SnapKit
MBProgressHUD
Loading and progress indicators
MIT License / https://github.com/jdg/MBProgressHUD
TZImagePickerController
Photo or media selection
MIT License / https://github.com/banchichen/TZImagePickerController
FFmpegKit / FFmpeg and bundled external libraries
Audio/video analysis, transcoding, audio extraction, audio/video composition
LGPL v3 and bundled license texts / https://github.com/arthenica/ffmpeg-kit, https://ffmpeg.org
Apple StoreKit
Subscriptions, in-app purchases, and entitlement status
Apple platform rules
Apple AVFoundation / Core ML and other system frameworks
Audio/video playback, recording, processing, and model inference
Apple platform rules
If crash analytics, advertising, statistics, cloud processing, account login, cloud storage, model downloads, or other third-party SDKs are added in the future, this notice and the Privacy Policy will be updated accordingly.
For open-source components that require retention of copyright notices, license texts, or disclaimers, the App will provide the corresponding notices in reasonable locations, such as:
"Settings - About - Third-Party Open-Source Notices" inside the App;
the "Third-Party Open-Source Components and Model Notice" page on the official website;
license files distributed with the App;
technical support pages.
If a third-party component's license differs from this notice, the original license text of that component shall prevail.
Third-party components, model weights, audio/video codecs, and copyrighted content processing may involve their respective license, copyright, patent, privacy, and platform rule requirements. When using the App to import, process, separate, export, compose, or share audio/video content, users should ensure that they have lawful rights to the relevant content or have obtained sufficient authorization.