- Analytic Solver Upgrade 可以最多求解最多2,000决策变量的线性问题、500 variables 决策变量的非线性问题，价格是995美元
- Analytic Solver Optimization 可以最多求解最多 8,000 决策变量的问题，价格是1995美元
- Solver engines 求解问题的规模几乎没有限制，不同版本价格存在差异，在3000到10795美元之间
OpenSolver是一个开源的工具、完全免费，但是求解问题的规模没有限制。OpenSolver由新西兰奥克兰大学工程科学系（Engineering Science department, University of Auckland）的Andrew Mason及学生开发、维护；最近的开发由麻省理工学院的Jack Dunn提供。
These solvers all handle linear optimization problems in both continuous and integer variables. Their performance is not at the level of analogous commercial solvers, but can be competitive for problems that are not too large or difficult.
The solvers in this category seek solutions to problems involving smooth nonlinear functions such as powers, logs, and ratios. They differ in the algorithms that they offer, and hence in their effectiveness for different problem types. Due to the difficulty of nonlinear optimization, these solvers are effective with smaller problems than their linear counterparts.
Ipopt 3.12.13 — from COIN-OR under the Eclipse Public License; available as source code and binaries for 32-bit Linux, 64-bit Linux, OS X, 32-bit Windows and 64-bit Windows. Finds locally optimal solutions to continuous nonlinear problems, using an interior-point method.
Bonmin 1.8.8 — from COIN-OR under the Eclipse Public License; available as source code and binaries for 32-bit Linux, 64-bit Linux, OS X, 32-bit Windows and 64-bit Windows. Finds globally optimal solutions to convex nonlinear problems in continuous and discrete variables, and may be applied heuristically to nonconvex problems.
Couenne 0.5.8 — from COIN-OR under the Eclipse Public License; available as source code and binaries for 32-bit Linux, 64-bit Linux, OS X, 32-bit Windows and 64-bit Windows. Finds globally optimal solutions to nonlinear problems in continuous and discrete variables, regardless of convexity.
Constraint programming solvers
These solvers handle constraint programming problems usually in discrete variables. They support a wide variety of constraint types that may contain nonlinear and logical expressions. Constraint programming solvers can be more efficient than MIP solvers for some kinds of combinatorial optimization problems.
The kestrel program allows using remote NEOS solvers with AMPL running on your local machine. Invoked in the same way as other AMPL solvers, Kestrel sends the problem to a solver running on one of the NEOS Server’s remote computers. The results from the NEOS Server are eventually returned through Kestrel to AMPL, where you can view and manipulate them locally in the usual way. Thus you get all the benefits of using AMPL environment, without having to first obtain and install each solver you want to try. Kestrel is provided free of charge and available for download from the Run AMPL on NEOS page. It is also included in all the demo packages.