---
id: gekko
version: "1.3.2"
license: MIT
license_treatment: permissive
maintenance: active
---

# gekko — Machine learning and optimization for dynamic systems

License: permissive · Maintenance: active · Downloads: 141.1K/mo

## What it is and what it does

GEKKO is an optimization and machine learning framework that translates mathematical models of dynamic systems into byte-code and solves them using large-scale solvers. It specializes in problems involving differential algebraic equations (DAEs)—systems where variables evolve over time according to differential equations and algebraic constraints. The package supports a wide range of problem formulations: from simple linear and quadratic programs to complex mixed-integer nonlinear problems, as well as control and estimation tasks like model predictive control and moving horizon estimation.

The package works by building a symbolic representation of your problem—defining variables, equations, and objectives—then compiling that representation and passing it to industrial-grade solvers (APOPT and IPOPT) with automatically computed sparse derivatives. It includes data-cleaning utilities and is designed to run on Windows, Linux, macOS, and ARM processors. GEKKO is intended for engineers and researchers solving real-world optimization and control problems, from parameter estimation and regression to real-time optimization and advanced process control.

Use it for:

- Solve differential algebraic equation systems with constraints and optimization objectives over a time horizon
- Build and tune model predictive control (MPC) systems for process automation or robotics
- Perform parameter estimation and data regression on dynamic system models
- Formulate and solve mixed-integer nonlinear programs (MINLP) for scheduling or resource allocation
- Implement moving horizon estimation (MHE) for state estimation in real-time systems
- Conduct sensitivity analysis and optimization of complex engineering systems

## Worth the install?

AI-flagged interpretation of the facts above — verify before relying.

GEKKO is a Python package for optimization and machine learning that specializes in solving dynamic systems described by differential algebraic equations, with support for linear, quadratic, nonlinear, and mixed-integer programming problems.

Yes. GEKKO is actively maintained, has no known vulnerabilities, installs with minimal friction (numpy only), and is licensed permissively. It fills a specialized niche in dynamic optimization and DAE solving that few Python packages address directly. Install if you need to solve differential algebraic equations, implement model predictive control, or tackle mixed-integer nonlinear optimization problems.

## Install

pip install gekko
uv add gekko
poetry add gekko

## Installing gekko

Before you install: Low install friction; pure Python wheel with only numpy as a runtime dependency. Active maintenance with recent commits and a stable production release status.

License in practice: MIT license permits unrestricted use, modification, and distribution with minimal restrictions—suitable for commercial and proprietary projects.

Quickstart:

pip install gekko
import gekko
m = gekko.GEKKO()
x = m.Var()
m.Equation(x**2 == 4)
m.solve()
print(x.value)

Verify before relying:

- Whether the bundled APOPT and IPOPT solvers are included in the wheel or require separate installation
- Performance characteristics and scalability limits for large-scale DAE systems
- Whether cloud/edge deployment modes require additional configuration or credentials

## Package facts

- License: MIT (permissive)
- Python support: supports_current
- Install friction: low
- Maintenance: active
- Downloads: 141.1K/month (top 15,000 on PyPI)
- Known vulnerabilities: none known

## Tags

differential algebraic equations solver, nonlinear optimization python, dynamic systems optimization, model predictive control, mixed-integer programming, machine learning optimization, DAE solver, optimization, control-systems, differential-equations

[View on SkillFed](https://skillfed.io/packages/gekko) · [View on PyPI](https://pypi.org/project/gekko/)