Open source licenses
This editor runs entirely in your browser, which means we send you the software rather than run it for you. Everything listed below is third-party work we ship or load, with the licence it is offered under and a link to its source. Licences here were read from each project's own metadata, not from memory.
GPL notice and written offer for source
The export feature is powered by FFmpeg n5.1.4 compiled to WebAssembly and distributed as @ffmpeg/core-mt 0.12.10. That build is configured with --enable-gpl against x264 and x265, so the resulting binary is licensed under the GNU General Public License, version 2 or later. We serve that binary to your browser unmodified, as /vendor/ffmpeg/ffmpeg-core.wasm. We have made no changes to it.
Written offer
The complete corresponding source code for the FFmpeg WebAssembly binary distributed by this site is publicly available at the links below. In addition, ReachUp LTD will, on request, provide the complete corresponding machine-readable source code for that binary on a medium customarily used for software interchange, for no more than our cost of physically performing the distribution. Write to apps@reachuphq.com. This offer is valid for three years from the date this version was distributed.
- Build definition that produces this exact binary: ffmpegwasm/ffmpeg.wasm @ v12.15
- FFmpeg source: FFmpeg @ n5.1.4
- Full licence text, served verbatim from this site: GNU General Public License v2.0
Shipped to your browser
Served from this origin as part of the application.
| Component | Version | Licence | Role | Source |
|---|---|---|---|---|
| @ffmpeg/core-mt | 0.12.10 | GPL-2.0-or-later | Multi-threaded FFmpeg WebAssembly build (ffmpeg-core.wasm) — audio extraction, trimming and export | Source |
| FFmpeg | n5.1.4 | GPL-2.0-or-later (as built) | Compiled into ffmpeg-core.wasm with --enable-gpl --enable-libx264 --enable-libx265 | Source |
| @ffmpeg/ffmpeg | 0.12.15 | MIT | JavaScript wrapper that drives the FFmpeg WebAssembly core | Source |
| @ffmpeg/util | 0.12.2 | MIT | Helper utilities for the FFmpeg wrapper | Source |
| onnxruntime-web | 1.26.0-dev.20260416-b7804b056c | MIT | WebAssembly / WebGPU inference backend for the speech models | Source |
| @huggingface/transformers | 4.2.0 | Apache-2.0 | Runs the speech models in the browser (transformers.js) | Source |
| coi-serviceworker | 0.1.7 | MIT | Cross-origin-isolation fallback so SharedArrayBuffer stays available | Source |
| Next.js | 16.2.12 | MIT | Application framework | Source |
| React / React DOM | 19.2.4 | MIT | UI runtime | Source |
| Tailwind CSS | 4.3.3 | MIT | Styling | Source |
| lucide-react | 1.27.0 | ISC | Icons | Source |
| zustand | 5.0.14 | MIT | Editor state store | Source |
| posthog-js | 1.407.5 | Apache-2.0 AND MIT | Self-hosted, anonymous product analytics | Source |
| Inter | served via next/font | SIL Open Font License 1.1 | Typeface, self-hosted (no external font request) | Source |
Compiled into the FFmpeg WebAssembly binary
These libraries are linked into ffmpeg-core.wasm rather than shipped separately. x264 and x265 are the reason the combined binary is GPL rather than LGPL. Versions come from the ffmpeg.wasm build definition; licences from each project's own COPYING file.
| Component | Version | Licence | Role | Source |
|---|---|---|---|---|
| x264 | branch 4-cores | GPL-2.0-or-later | H.264 encoder | Source |
| x265 | 3.4 | GPL-2.0-or-later | H.265 encoder | Source |
| LAME | master | LGPL-2.0-or-later | MP3 encoder | Source |
| libvpx | v1.13.1 | BSD-3-Clause (WebM Project) | VP8 / VP9 codec | Source |
| Ogg / Vorbis / Theora / Opus | 1.3.4 / 1.1.1 / 1.3.1 | BSD-3-Clause (Xiph.Org) | Container and audio/video codecs | Source |
Speech models
Model weights are downloaded by your browser directly from the Hugging Face CDN the first time you transcribe, and cached there. We never host or redistribute them.
| Component | Version | Licence | Role | Source |
|---|---|---|---|---|
| onnx-community/whisper-base_timestamped | ONNX export of openai/whisper-base | Apache-2.0 | Speech-to-text with per-word timestamps | Source |
| onnx-community/whisper-small_timestamped | ONNX export of openai/whisper-small | Apache-2.0 | Speech-to-text, higher accuracy | Source |
| onnx-community/pyannote-segmentation-3.0 | ONNX export of pyannote/segmentation-3.0 | MIT | Speaker segmentation and diarization | Source |
| onnx-community/silero-vad | ONNX export of Silero VAD | MIT | Voice-activity detection | Source |
Upstream project
This editor is a modified fork of Rescript by Wassim Gharbi (@wassgha), used under the MIT License. We rebranded it, removed Google Analytics, changed the hosting model and reworked the interface. The upstream copyright notice and the full MIT permission notice are preserved in the LICENSE and NOTICE.md files that travel with the source.
Spotted something wrong or missing on this page? Tell us at apps@reachuphq.com and we will correct it.