Pedalboard

Audio processing at C++ speed from Python is what Pedalboard delivers: Spotify's library wraps the JUCE framework to provide studio-quality effects, reading and writing of WAV, AIFF, FLAC, MP3, and OGG files, and hosting of third-party VST3 and Audio Unit plugins inside ordinary Python code. Built-in processors include compressor, chorus, distortion, reverb, phaser, pitch shift, limiter, and convolution, chainable into a Pedalboard object that renders far faster than real time, while an AudioStream class handles live input and output. Spotify built it for machine learning data augmentation, and it feeds TensorFlow tf.data pipelines directly, but podcasters and producers also use it for batch processing and effect rendering. Wheels for Linux, macOS including Apple Silicon, and Windows install with pip install pedalboard on Python 3.10 through 3.14, and no GPU is involved. The license is GPL-3.0, inherited from JUCE, which matters for closed-source redistribution though not for internal use. About 6,200 GitHub stars and production use inside Spotify's AI DJ and voice translation features anchor its reputation.