Solution Guarantees and Best-Bound Access for Convex MINLPs in Gurobi via AMPL

Answered

Niloufar Sadeghi

December 13, 2025 18:37

• 

Hello,

I’m using Gurobi to solve a convex mixed-integer nonlinear problem (one constraint includes an exponential term). I model it in AMPL and call the solver from Python via amplpy.

1. With Gurobi 13, can I be confident that my problem is solved exactly?
2. If I set a time limit, how can I retrieve the best bound found so far?

Thank you.

Related to

• AMPL

0

• 

  Riley Clement

  • Gurobi Staff

  December 16, 2025 04:05

  Hi Niloufar,

  With Gurobi 13, can I be confident that my problem is solved exactly?

  In theory, given enough time and memory, and a well-conditioned model, then Gurobi will solve to optimality (as global solvers are intended to do).

  I model it in AMPL and call the solver from Python via amplpy …. If I set a time limit, how can I retrieve the best bound found so far?

  I had a quick look at the amplpy docs and it's not obvious how to do this.  This question is probably best posed to people who use AMPL: https://discuss.ampl.com/latest

  - Riley

  0

Please sign in to leave a comment.