SliceStep is built on the work of others. This page lists the third-party software and models distributed with, or downloaded by, the application, along with their licenses and any required notices.
Used for: decoding the screen recording and generating video, GIF, and image output — entirely on your device.
License: GNU General Public License, version 3 (GPL v3). The bundled build is configured with --enable-gpl --enable-version3 and includes GPL-licensed components such as libx264.
Version bundled: [FILL IN — the exact version string from the bundled binary, shown by running ffmpeg.exe -version. Record it verbatim; the source offer below must correspond to this exact build.]
ffmpeg is bundled as a standalone executable and is invoked by SliceStep as a separate program via its command-line interface. It is not linked into, combined with, or modified by SliceStep. SliceStep itself is proprietary software and is not derived from ffmpeg; only the ffmpeg executable is covered by the GPL.
Corresponding source: The complete source code for the exact ffmpeg build distributed with SliceStep is available. You may obtain it from the FFmpeg project at ffmpeg.org and from the build publisher's source and build scripts at gyan.dev/ffmpeg/builds. In addition, we will provide the corresponding source for the exact bundled version on request, at no charge beyond the cost of distribution, for at least three years — contact hello@slicestep.com.
The full text of the GPL v3 is reproduced here: [PASTE FULL GPL v3 TEXT, or link to a copy hosted on this site. The same text ships with the app at build/licenses/ffmpeg-LICENSE.txt.]
Used for: optional, off-by-default AI suggestions for instruction text, and on-screen text recognition used by the redaction feature.
License: MIT License.
Source: Microsoft, distributed via Hugging Face (onnx-community/Florence-2-base-ft and onnx-community/Florence-2-large-ft).
Model files are downloaded only if you enable AI features, and run entirely on your device thereafter.
Used for: on-device transcription of voice narration.
License: MIT License.
Source: OpenAI, distributed via Hugging Face. [FILL IN — exact model variant bundled or downloaded]
Used for: running the above models locally in the application.
License: Apache License 2.0.
Source: Hugging Face (@huggingface/transformers).
Used for: the desktop application framework.
License: MIT License. Electron includes Chromium (BSD-style licenses) and Node.js (MIT), each under their own terms.
[FILL IN — list remaining bundled npm packages and their licenses, e.g. Jimp (MIT). A generated license report from your build tooling is the most reliable source for this list.]
If you need a copy of the corresponding source for ffmpeg or any other component, or have a licensing question, contact hello@slicestep.com.
