The Echo Nest Analyze API is a tool that can enrich your software’s understanding of musical recordings. It uses a perceptual model of human listening to generate detailed descriptions of a song’s structure and musical content. Given any musical sound file, the analysis engine outputs global musical track information such as tempo, dominant pitches and timbre, key, time signature, energy, etc. It also segments the sound file into large sections (such as verse, chorus, bridge), beats, and smaller time slices. Each segment is described in terms of pitch, timbre, and loudness content. The engine then segments a musical track into small time slices, each representing a perceived “sound” (a note of a piano, a syllable of a vocal line, etc.). Each segment is then described in pitch, timbre and loudness. This information can power performances, visualizations, and games with the ability to respond to a wide range of nuances in the music they use.

Developers can freely access the current analysis via a web services API that, given an MP3, will return an XML file that describes all the information we can extract from a song file. Visit the registration page to request an API key. For more direct and commercial use, they also offer heavily optimized libraries and stand alone binaries for Linux, Mac OS X, and Windows platforms on Intel, AMD, Power PC, and ARM hardware. Please contact audioapi@echonest.com for more information on direct licensing.

The Echo Nest Analyze API powers the automatic beat matching technology “Capsule.” As we know the beat of a track at any moment, we can seamlessly transition between two different pieces of music in real time. You can try this technology yourself on This is my jam at http://www.thisismyjam.com/