Skip to the content
Know your tech before you buy it. Makers: get listed from $49/mo

music21

music21.org

  • Free plan
  • Windows
  • macOS
  • Linux
  • Web
8.2 Club score Visit music21.org

What music21 is

music21 is a free, Python-based toolkit for computer-aided musicology. It supports work such as analyzing music, studying datasets, creating musical examples, teaching music theory, editing notation, and composing. Its corpus contains thousands of encoded musical works, while its analysis tools include corpus searching, pitch plots, and functions such as nPVI. The toolkit handles common formats including MIDI and MusicXML and can be extended to support more. For notation work, its documentation recommends MuseScore and describes MusicXML readers including Finale, Sibelius, and Dorico. Optional integrations include LilyPond, Matplotlib, SciPy, PyAudio, and Pygame for uses such as score rendering, plotting, audio searching, recording, and realtime MIDI. It can also be used in web-based Jupyter Notebook. music21 is BSD-licensed, but some corpus works have separate commercial-use or other restrictions, so their individual licenses matter. Users need some familiarity with Python. The toolkit primarily works with symbolic musical data rather than sound waves and cannot read notation directly from images or PDFs.

Who it suits

music21 is suited to musicologists, programmers, composers, performers, and music enthusiasts who have some Python familiarity. It is for symbolic music work, not reading notation from images or PDFs.

Good to know, and what to watch

  • Includes a corpus of thousands of encoded works.
  • Supports MIDI and MusicXML.
  • Offers corpus searching and pitch plotting.
  • Can be used in web-based Jupyter Notebook.
  • Requires some familiarity with Python.
  • Focuses on symbolic data rather than sound waves.
  • Cannot read notation directly from images or PDFs.
  • Some corpus works have separate restrictions.

music21

Free

Free plan

Getting started

  1. Open the music21 website
  2. Choose a Linux, macOS, Windows, or web environment
  3. Use Python to work with the toolkit
  4. Load MIDI or MusicXML material, or search the included corpus
  5. Open notebooks through the web-based Jupyter Notebook
  6. Add optional tools such as MuseScore or LilyPond for notation work

Limits to check

The toolkit does not process sound waves as its primary data type and cannot read notation directly from images or PDFs. Some corpus encodings have commercial or other licensing restrictions that must be checked for individual works.

The Club's review

music21 is a Python toolkit for computer-aided musicology, built for working with encoded scores through code. It suits Python-familiar researchers, teachers, composers and technically minded performers; its main payoff is analysis and transformation of symbolic music, not a finished notation or audio-transcription workflow.

Overview

music21 combines a corpus of thousands of encoded musical works with Python tools for searching, editing, generating and studying them. That makes it possible to investigate a question across many scores, produce teaching examples, or prototype compositional ideas in a repeatable way. The approach rewards users comfortable with programming; people seeking a visual, self-contained score editor will find its center of gravity elsewhere.

The toolkit is BSD-licensed and free. Corpus licensing is a separate practical concern: individual works or directories can restrict commercial use or impose other conditions, so check the terms of each item before including it in a commercial project.

Key features

Corpus search and analysis

Corpus searching, pitch plotting and advanced measures such as nPVI support data-led musicology and classroom analysis. Instead of inspecting one score at a time, users can compare encoded material and turn findings into generated examples. The included corpus gives substantial material to explore, although its mixed licensing means it is not automatically cleared for every use.

Formats and score manipulation

music21 reads and writes common symbolic formats including MIDI and MusicXML, and its format model can be extended. Parts extraction is useful when a full score needs to be separated for study or preparation, while tablature support broadens its reach beyond conventional staff notation. It manipulates symbolic representations rather than sound waves, so it is not a tool for analyzing recordings.

Notation and connected tools

For viewing and editing notation, the documentation recommends MuseScore, with MusicXML readers including Finale, Sibelius and Dorico also supported. Optional integrations with LilyPond, Matplotlib, SciPy, PyAudio and Pygame extend the workflow to rendering, plotting, audio searching, recording and realtime MIDI. Those integrations add useful routes for specialist projects, but they also mean users may need to assemble and configure a broader toolchain.

Notebook workflow

music21 can run in the web-based Jupyter Notebook, giving educators and analysts an interactive place to combine code, explanation and musical output. Its configuration assistant asks about Internet access, but Internet access is not required. Community help is available through a mailing list, and paid consulting is offered at project-dependent rates.

Pricing

PlanPriceWhat it includes
music21FreeBSD-licensed toolkit; corpus works may have separate restrictions

There are no paid tiers, seat counts or renewal terms to weigh. The trade-off for the zero price is that users supply the Python knowledge and surrounding notation or analysis applications needed for their workflow; corpus permissions must still be checked work by work.

Platforms

music21 is available on Linux, macOS, Windows and the web, with both local and hosted deployment options. The web route is particularly relevant to Jupyter-based work, while local deployment suits projects that need a controlled programming environment.

Who it's for

Choose music21 if you are a musicologist searching and measuring encoded scores, a teacher building computational theory examples, or a composer exploring algorithmic ideas. Performers and enthusiasts can benefit when they are willing to work in Python. It is a poor fit if your priority is editing notation visually without code, converting images or PDFs into readable scores, or processing audio directly.

Pros and cons

  • Pro: A large encoded corpus and search functions support investigations that would be laborious by manual score inspection.
  • Pro: MIDI, MusicXML, parts extraction and tablature cover varied symbolic-music workflows.
  • Pro: BSD licensing and a free plan remove software licensing costs.
  • Con: Python familiarity is a stated requirement, which raises the entry barrier for non-programmers.
  • Con: It works with symbolic data, not sound waves, and cannot read notation directly from images or PDFs.
  • Con: Corpus works can have separate commercial or other restrictions, complicating redistribution and commercial projects.

Alternatives

For a broader starting point, browse Music Theory Software or Music notation software. Choose Harmonia instead when guided practice and feedback across mobile, desktop and web matter more than programmable corpus analysis; its free practice tier omits advanced analyzing and feedback features, while Classroom Connect costs 20.95 for up to four weeks. note-noodle is a better direction for ear training and play-along exercises, with a free daily and monthly allowance. ToneSavvy suits web-based teaching for groups, offering a plan for up to 25 students at $19 a month. MyMusicTheory provides more than 100 free exercises for conventional theory practice. MusicTheory.net offers its full, ad-free website content at no cost on iOS and the web. Musition, ChordIQ and Complete Rhythm Trainer are other alternatives for structured theory or rhythm learning.

Verdict

music21 is the right choice for Python-capable users who need a free, extensible environment for searching, analyzing and generating symbolic music across a substantial corpus. Its strongest reason to choose it is the combination of computational analysis and broad format support. Look elsewhere if you need visual-first notation editing, audio analysis, or reliable permission to use every corpus work commercially.

The verdict

music21 offers a broad set of tools for computational work with symbolic music, with formats and integrations for notation and analysis. Users should check corpus licenses and note that it does not interpret notation from images or PDFs.

Questions people ask

How much does music21 cost?

The music21 plan is free at no cost, and the toolkit is BSD-licensed.

Which platforms support music21?

It supports Linux, macOS, Windows, and the web, with both local and web deployment.

Which file formats can it use?

It works with MIDI and MusicXML and can be extended to support additional formats.

Can music21 edit notation?

Yes. Its uses include editing notation, and documentation recommends MuseScore plus MusicXML readers such as Finale, Sibelius, and Dorico.

Does it work with Jupyter?

Yes. The documentation describes using music21 in the web-based Jupyter Notebook.

Is the corpus unrestricted for commercial use?

Some corpus works may have commercial or other restrictions, so each work or directory's license should be checked.

What the maker publishes

Analysis
Its tools include corpus searching, pitch plotting, and advanced analysis functions such as nPVI. source
Audience and requirements
The documentation says users need some familiarity with Python and identifies musicologists, programmers, composers, performers, and music enthusiasts among potential contributors. source
Corpus
The toolkit includes a corpus of thousands of encoded musical works. source
Corpus restrictions
Some corpus encodings may not be used commercially or may have other restrictions, so users are directed to check the licenses for individual works or directories. source
Deployment
both source
Formats
The FAQ says music21 can work with common formats including MIDI and MusicXML, and can be extended to support more. source
Free plan
Yes source
Jupyter
The documentation describes using music21 in the web-based Jupyter Notebook. source
Limits
music21 primarily manipulates symbolic musical data rather than sound waves, and it cannot read music directly from images or PDFs. source
MIDI support
Yes source
MusicXML support
Yes source
Notation integrations
The documentation recommends MuseScore for viewing and editing notation and describes support for MusicXML readers including Finale, Sibelius, and Dorico. source
Other integrations
The documentation describes optional integrations with LilyPond, Matplotlib, SciPy, PyAudio, and Pygame for tasks such as score rendering, plotting, audio searching, recording, and realtime MIDI. source
Parts extraction
Yes source
Purpose
music21 is a Python-based toolkit for computer-aided musicology. source
Security and licensing
The toolkit is BSD-licensed; its configuration assistant asks whether music21 can access the Internet, and the documentation says Internet access is not required. source
Support
The community mailing list offers help, while paid consulting is available at rates that vary by project. source
Tablature support
Yes source
Uses
People use it to study music datasets, generate musical examples, teach music theory, edit notation, analyze music, and compose. source

Checked against the maker's pages on 8 Oct 2026.