openSMILE
audeering.com
- Free plan
- Windows
- macOS
- Linux
- iPhone and iPad
- Android
- API
- Self-hosted
8.0 Club score Visit audeering.com
Plans
Every plan and priceCommercial development license
—
openSMILE Research License
Free
Open-source (private, research, and educational)
Free
Free plan
What the maker publishes
- Adoption
- audEERING reports more than 150,000 downloads since the 2010 publication and more than 2,650 citations in academic papers. source
- API wrappers
- openSMILE includes a C API with wrappers for Python, .NET, Android, and iOS. source
- Audience
- It is intended primarily for research applications, demonstrators, and prototypes, and its target users are researchers and system developers. source
- Audio integrations
- Audio input options include FFmpeg-supported media files, PortAudio live recording and playback, and Android/iOS recording APIs. source
- Binary input limit
- Official binaries are not built with FFmpeg, and default configs use the WAVE source component. source
- Commercial licensing
- audEERING directs users seeking commercial openSMILE use to contact it for a separate license. source
- Commercial limit
- The open-source version may not be used for any sort of commercial product. source
- Feature sets
- The Python package supports ComParE 2016, GeMAPS, and eGeMAPS feature sets. source
- Implementation
- It is written in C++ and designed for desktop, mobile, and embedded platforms. source
- License boundary
- The research license permits non-commercial purposes; other commercial use requires a commercial license or written approval. source
- Maker
- audEERING GmbH is headquartered in Gilching, Germany, and its company profile says it was founded in 2012. source
- Notable limit
- The openSMILE Python package supports only 64-bit Python. source
- Output formats
- Feature output/read formats include CSV, WEKA ARFF, HTK parameter files, and LibSVM output. source
- Platforms
- It runs on desktop, mobile, and embedded platforms including Linux, Windows, macOS, Android, iOS, and Raspberry Pi. source
- Processing capabilities
- The toolkit supports incremental real-time processing, modular reusable components, and plugins. source
- Processing mode
- openSMILE is designed for real-time online processing and can also run offline in batch mode on large data sets. source
- Purpose
- openSMILE is an open-source toolkit for audio feature extraction and classification of speech and music signals. source
- Python interface
- A standalone Python package provides an interface for extracting openSMILE features and can be installed with pip. source
- Python package
- The opensmile Python package extracts features from inside Python, and a separate openSMILE installation is not required. source
- Support
- Feedback and technical support are requested through the GitHub issue tracker. source
- Support
- The project recommends GitHub as the place to discuss openSMILE, seek help, and report or track issues. source
- Threading limit
- Multi-threaded processing is supported only in versions 2.3 and lower; from version 3.0, processing always uses a single thread. source
- Throughput
- The product page states that 27k features can be extracted with a real-time factor of 0.08. source
- Version 3.0 speed
- audEERING reports a consistent 40–50% reduction in processing time for commonly used feature sets versus openSMILE 2.3. source
Checked against the maker's pages on 14 Nov 2026.