Skip to content

Latest commit

 

History

History
71 lines (51 loc) · 2.61 KB

File metadata and controls

71 lines (51 loc) · 2.61 KB

Third-Party Notices

This document summarizes third-party components used by VibeMouse and their declared licenses.

Last reviewed: 2026-03-01.

1) Project License

VibeMouse source code is licensed under Apache-2.0. See LICENSE.

2) Direct Python Dependencies

The following are direct runtime dependencies declared in pyproject.toml.

Package Declared License Upstream
numpy BSD-3-Clause https://numpy.org
sounddevice MIT https://python-sounddevice.readthedocs.io/
soundfile BSD-3-Clause https://github.com/bastibe/python-soundfile
pynput LGPL-3.0 (or later) https://github.com/moses-palmer/pynput
evdev BSD-3-Clause https://github.com/gvalkov/python-evdev
PyGObject LGPL-2.1 (or later) https://pygobject.gnome.org
pyperclip BSD https://github.com/asweigart/pyperclip
funasr MIT https://github.com/modelscope/FunASR
funasr-onnx MIT https://pypi.org/project/funasr-onnx/
onnxruntime MIT https://github.com/microsoft/onnxruntime
openvino Apache-2.0 https://github.com/openvinotoolkit/openvino
modelscope Apache-2.0 https://github.com/modelscope/modelscope

Notes:

  • pynput and PyGObject are LGPL-licensed. If you redistribute packaged binaries, ensure LGPL obligations are satisfied (license notice, relinking conditions where applicable, and source availability requirements for the LGPL-covered components).
  • Transitive dependencies are not exhaustively listed here. They remain subject to their own licenses.

3) Model Assets and Weights

VibeMouse defaults to model IDs:

  • iic/SenseVoiceSmall
  • iic/SenseVoiceSmall-onnx

At review time, ModelScope API metadata reports both model IDs as Apache License 2.0.

References:

Important caveat:

4) Attribution and Compliance Guidance

When distributing VibeMouse (source or binaries):

  1. Keep LICENSE and this notice file.
  2. Preserve upstream copyright and license notices for bundled components.
  3. Re-check model card licenses when changing model IDs or revisions.
  4. Do not assume all speech models under the same ecosystem share identical license terms.