DOWNLOAD A FREE DEMO

Get started quickly by downloading a package of AMPL and solvers from this page. Use your package in either of these ways:

• As a size-limited demonstration version, for getting a fast start on learning what AMPL is like and how it works. Install and use this version immediately to run all of the examples in the freely available AMPL book, which provides a tutorial in both optimization modeling and the AMPL system. Also use this version to experiment with building your own models and solving small instances that fit within the size limitations (listed below).

• As a full-featured 30-day trial version, for experimenting with AMPL and solvers on your optimization applications without any limitations as to problem size. To request a trial license file for this purpose, simply fill out and submit the short online form on our Request a Full Trial page (except if you are a student in which case you must use the form on our AMPL for Students page instead).

To download your package of AMPL and solvers, follow the instructions below for your preferred platform:
Windows, Linux, or macOS.

Demo Package Details

Demonstration packages are available for your choice of AMPL modeling environments:

• The AMPL IDE, an integrated command processor and editor that assists in model development.
• Command-line AMPL, which runs in a simple command window.

Your package includes a free license that permits unrestricted use for evaluation or study, subject to limits on the size of problem that can be sent to a solver:

• For linear problems, 500 variables and 500 constraints plus objectives.
• For nonlinear problems, 300 variables and 300 constraints plus objectives.

These limitations are applied after any presolve reductions and elimination of variables. The following solvers (some with stricter limits as shown) are included in the demo packages:

• CPLEX, Gurobi, and Xpress for linear and quadratic problems in continuous and integer variables.
• KNITRO for smooth nonlinear problems in continuous and integer variables (limit 10).
• CONOPT, LOQO, MINOS, and SNOPT for linear and smooth nonlinear problems in continuous variables.
• BARON for global nonlinear optimization in continuous and integer variables (limit 10).
• ILOG CP for a constraint programming approach to discrete variables and diverse logical conditions.
• LGO for problems that have many locally optimal solutions or nonsmooth functional forms (limit 10).
  

Consult the Solvers We Sell page or follow the individual links above for more solver information. (Separately you can also download the AMPL versions of popular free solvers from our Open Source Solvers page.)

To get started, follow the download instructions below for Windows, Linux, or macOS.

Alternatives for Trying AMPL

The AMPL demo version is ideal for getting a quick start on learning what AMPL is like and how it works. Subject to the problem size limits given above, it supports all AMPL modeling features; it is compatible with all of the examples in the AMPL book, which is also freely downloadable. When the time comes to try AMPL on larger problems than the demo version will permit, consider one of the following options, which are not subject to size limitations:

• Your demo download can be converted to a free 30-day trial by filling out the form on our Request a Full Trial page (or for students, our AMPL for Students page). As soon as your request is approved, we will email you a trial license file to replace the demo license that came with your download.

• AMPL for Courses provides packages of AMPL plus solvers that can be given out to students and used for any purpose in a designated course. The packages time out after the course is over.

• The NEOS Server is a free “cloud” optimization service that makes AMPL and most solvers available for trial through an online scheduler and a collection of powerful workstations.

For large-scale research and commercial projects you can purchase AMPL and solvers under a variety of licensing arrangements.

AMPL IDE download for Windows

To install: Download the distribution zipfile, amplide.mswin32.zip for the 32-bit version or amplide.mswin64.zip for the 64-bit version. Double-click the zipfile icon, or apply an unzip utility, to extract the folder named amplide.mswin32 or amplide.mswin64 from the zipfile. This will be your AMPL folder; optionally you may rename it and you may move it to any convenient location on your computer.

To run: Inside your AMPL folder, double-click the amplide folder icon to open that folder, and then double-click the amplide.exe file icon to start the AMPL IDE application. A small “AMPL IDE” window will appear while the program is being loaded, and then the full IDE application window will open. To get started using AMPL IDE, choose Help Contents from the Help menu at the top of the application window.

AMPL Command Line download for Windows

To install: Download the distribution zipfile, ampl.mswin32.zip for the 32-bit version or ampl.mswin64.zip for the 64-bit version. Double-click the zipfile icon, or apply an unzip utility, to extract the folder named ampl.mswin32 or ampl.mswin64 from the zipfile. This will be your AMPL folder; optionally you may rename it and you may move it to any convenient location on your computer.

To run: Inside your AMPL folder, double-click the sw.exe icon and type ampl at the prompt in the window that appears. Then you will see an ampl: prompt and can proceed to type AMPL commands. By default, AMPL model and data statements will refer to files that you have saved in your AMPL folder; you can instead read from the models folder of AMPL book examples (which is part of the installation) by giving the command option ampl_include models;. Be sure to choose a solver, by giving a command of the form option solver xxx; with xxx replaced by the name of the solver that you wish to run.

AMPL IDE download for Linux

To install: Download one of the distribution archives, amplide.linux32.tgz for a 32-bit version or amplide.linux64.tgz for a 64-bit version. Then extract the contents of this archive by typing the command

tar xzf amplide.linux32.tgz

or

tar xzf amplide.linux64.tgz

When the extraction is complete you will see a directory named amplide.linux32 or amplide.linux64. This will be your AMPL folder; optionally you may rename it and you may move it to any convenient location on your computer.

To run: Inside your AMPL directory, you will find a directory named amplide. Use the cd command to make amplide your current directory, and then start AMPL IDE with the command ./amplide. A small “AMPL IDE” window will appear while the program is being loaded, and then the full IDE application window will open. To get started using AMPL IDE, choose Help Contents from the Help menu at the top of the application window.

AMPL Command Line download for Linux

To install: Download one of the distribution archives, ampl.linux32.tgz for a 32-bit version or ampl.linux64.tgz for a 64-bit version. Then extract the contents of this archive by typing the command

tar xzf ampl.linux32.tgz

or

tar xzf ampl.linux64.tgz

When the extraction is complete you will see a directory named ampl.linux32 or ampl.linux64. This will be your AMPL folder; optionally you may rename it and you may move it to any convenient location on your computer.

To run: In a command window, use cd to go to your AMPL directory, and type ./ampl at the system prompt. Then you will see an ampl: prompt and can proceed to type AMPL commands. By default, AMPL model and data statements will refer to files that you have saved in your AMPL folder; you can instead read from the models folder of AMPL book examples (which is part of the installation) by giving the command option ampl_include models;. Be sure to choose a solver, by giving a command of the form option solver ./xxx; with xxx replaced by the name of the solver that you wish to run.

AMPL IDE download for macOS

To install: Download the distribution archive file, amplide.macosx64.tgz. Double-click this file’s icon to extract its contents. When the extraction is complete you will see a folder named amplide.macosx64. This will be your AMPL folder; optionally you may rename it and you may move it to any convenient location on your computer.

To run: First, follow these steps which will confirm that you want to allow this application to run on your Mac: (1) In your amplide.macosx64 folder, find the Amplide file (with a black cat’s head icon). (2) Drag the Amplide file to your desktop. (3) Drag the Amplide file back into the amplide.macosx64 folder.

Second, while holding down the control key, click on Amplide. A small menu will open. In this menu, select Open. A box asking “Are you sure you want to open it” will appear. Click Open in that box.

Third, a small “AMPL IDE” window will appear while the program is being loaded, after which the full IDE application window will open. Now you can begin using AMPL. To get started with the AMPL IDE interface, choose Help Contents from the Help menu at the top of the application window.

The above setup steps are only needed once; after that, you can open the AMPL IDE by just double-clicking Amplide.

AMPL Command Line download for Mac OS X

To install: Download the distribution archive file, ampl.macosx64.tgz. Double-click this file’s icon to extract its contents. When the extraction is complete you will see a folder named ampl.macosx64. This will be your AMPL folder; optionally you may rename it and you may move it to any convenient location on your computer.

To run: Click on the magnifying glass at the top right of your screen, enter terminal in the Spotlight window, and click on the Terminal icon to open a Mac Terminal window. Use cd to go to your AMPL folder; for example if your AMPL folder is ampl.macosx64 on your desktop, type cd Desktop/ampl.macosx64. Then type ./ampl; you will see an ampl: prompt and can proceed to type AMPL commands. By default, AMPL model and data statements will refer to files that you have saved in your AMPL folder; you can instead read from the models folder of AMPL book examples (which is part of the installation) by giving the command option ampl_include models;. Be sure to choose a solver, by giving a command of the form option solver ./xxx; with xxx replaced by the name of the solver that you wish to run.