GNU Octave

GNU Octave is a free scientific programming language for numerical work. The home page puts it under the GNU General Public License and the project calls its syntax largely compatible with MATLAB, describing it as drop-in compatible with many MATLAB scripts. It runs on GNU/Linux, macOS, BSD and Microsoft Windows and includes built-in two and three dimensional plotting.

Interfaces include a GUI mode, a console and shell scripts, and further functions arrive as packages listed at packages.octave.org. The about page names John W. Eaton as copyright holder from 1998 to 2026 and states no country or organisation behind the project. The pages list no price; the only money-related link on the download page is Donate. The 11.3.0 manual covers linear algebra, optimisation, statistics, signal and image processing, differential equations and more.

• MATLAB-compatible syntax. The project calls the syntax largely compatible with MATLAB and says many scripts run as they are. That eases a move from MATLAB, though the pages do not promise full compatibility or list the functions that differ.
• Linear algebra on matrices. Vectors and matrices are the core data types, and the manual has chapters on linear algebra and sparse matrices. The pages give no numeric size limits, only a 64-bit indexing build option for very large arrays.
• Plotting in 2D and 3D. High-level commands draw two and three dimensional plots, and the home page lists plotting among the headline features. Plot output formats and rendering back ends are not detailed in the pages we read.
• Numerical method chapters. The 11.3.0 manual includes differential equations, optimisation, statistics, signal processing, image processing and audio processing. This spread suits general engineering and science scripting more than any single specialist field.
• Package system. Extra functions come as packages found at packages.octave.org. The manual describes installing, removing and creating them with a DESCRIPTION file, an INDEX file and PKG_ADD and PKG_DEL directives.
• Several ways to run code. Octave can be used through a GUI, a console or shell scripts. A GUI Development chapter in the manual covers dialogs, progress bars and other interface elements for those who build small tools.
• Extension in compiled code. Oct-Files and Mex-Files are described as mechanisms for compiled code, and an appendix covers a Java interface. The manual names no Python interface, which matters for teams that mix languages.

• A student replacing a MATLAB course licence. The MATLAB-like syntax means course scripts can often run as written, and the software carries no listed price. The pages do not say how well a particular toolbox function is covered.
• An engineer scripting numerical analysis. Linear algebra, optimisation, differential equations and signal processing chapters give the manual enough breadth for routine analysis. Compiled extensions are possible through Oct-Files or Mex-Files when a loop is too slow.
• A research group wanting source access. The source is a tarball on the GNU FTP mirror with GPG signature verification described, and the repository under the gnu-octave organisation shows GPL-3.0. That helps groups that must archive or audit what they run.

The download page offers a source tarball from the GNU FTP mirror and volunteer-made packages for GNU/Linux and for FreeBSD or OpenBSD. Windows users get a 64-bit installer of about 540 MB, plus a 64-bit linear algebra build that the page says offers no advantage unless the machine has roughly 32 GB of memory or more.

The 32-bit Windows build is no longer supported and its link has been removed. For macOS there is no direct installer: the page points to the Octave Wiki and to the third-party package managers Homebrew and MacPorts. Once installed, the manual at docs.octave.org is the reference for the GUI, console and packages.

GNU Octave is a free scientific programming language whose syntax the project calls largely compatible with MATLAB, so many scripts run unchanged. The open source code is GPL-3.0, and the manual covers linear algebra, optimisation and signal processing.

The gaps are practical: the project offers no direct macOS installer, the Windows installer is about 540 MB, and the manual documents a Java interface but names no Python one.

GNU Octave is entry #1 of 13 in the numerical computing software list: the next place goes to Julia. Places 1 to 3 in every list go to challengers where the category has them, so this is a position on that rule and not a score.

The vendor states no country of establishment, so GNU Octave is left out of the regional count for this list.

Charging model: Free software under the GNU GPL; the pages list no price, only a donate link.

The home page describes GNU Octave as free software under the GNU General Public License and gives no price. The download page also states no price, and its only money-related link is Donate, so the project appears to rely on voluntary contributions. There are no tiers, seats or licence keys described in the pages we read. The cost that does exist is practical: the Windows installer is about 540 MB, and the macOS route through Homebrew or MacPorts means relying on third-party package managers. The website content carries a separate Creative Commons Attribution-NoDerivatives 4.0 licence, which is distinct from the software licence. Paid support or commercial contracts are not described.

Source: the pricing material on octave.org, as read for Best Numerical Computing Software in 2026 and last reviewed on 13 November 2026.

The manual chapters cover plotting, sparse matrices, differential equations, optimisation, statistics and audio processing, and a packages chapter explains how to build an add-on with a DESCRIPTION file and an INDEX file.

Compiled code enters through Oct-Files and Mex-Files.

A 64-bit linear algebra build exists for Windows, though the page says it only pays off with roughly 32GB of memory.

Test each MATLAB script before relying on the compatibility claim.

The pages we read do not describe a hosted service, so data stays wherever the user installs and runs Octave. The project publishes its source as a tarball on the GNU FTP mirror and describes GPG signature verification for it on the download page. The repository under the gnu-octave organisation on GitHub shows the GPL-3.0 licence.

The about page names John W. Eaton as copyright holder and names no country or organisation as the place the project is run from. The manual keeps a list of known bugs that have not been fixed yet. Certifications and formal security statements are not listed.

Integrations the vendor names (10):

• MATLAB scripts
• Java
• Oct-Files
• Mex-Files
• packages.octave.org
• Homebrew
• MacPorts
• GNU FTP mirror
• GitHub
• GUI mode

The choice below rests on shared catalogue tags, price model, free plan and region, with rank distance as the tie-break.

• Julia (#2 of 13): Not for: teams wanting a bundled commercial toolbox with one vendor behind it. Fast dynamic language for scientific computing, released under the MIT licence.
• SMath Studio (#3 of 13): Units-aware maths worksheet program with published prices from SMath LLC.
• R Project (#4 of 13): A good match for: Statisticians and analysts who work mainly with models, tests and time series.

Read from the vendor’s own pages on 12 November 2026.

• octave.org: licence, platforms, syntax compatibility, features, no price
• octave.org/download.html: installers, macOS routes, 32-bit removal, GPG, donate link
• octave.org/about.html: copyright holder, website licence, no country stated
• docs.octave.org/latest: manual chapters, packages, Java interface, Oct and Mex files, bugs
• github.com/gnu-octave/octave: GPL-3.0 licence

Feature lists tell little, so positions rest on setup effort, the true price, portability and fit. On those tests GNU Octave lands at place 1 of 13. Method and commercial relationships are in our editorial process.