Gurobi Optimizer
gurobi.com
- Free plan
- Windows
- macOS
- Linux
- API
- Self-hosted
What Gurobi Optimizer is
Gurobi Optimizer turns mathematical models of business problems into solutions, supporting linear, mixed integer, quadratic, combinatorial, nonlinear, conic, and multi-objective problems. For non-convex quadratic problems, its global optimization can establish a global optimum rather than only a local one. Solver capabilities include parallel and distributed optimization, presolve, cutting planes, heuristics, and symmetry handling. For reviewing models and results, it provides sensitivity and infeasibility analysis, feasibility relaxation, and solution quality attributes. APIs are available for C, C++, Java, .NET, Python, MATLAB, and R, supporting model integration into applications, services, and data pipelines. Deployment choices include local, server, cloud, and container environments. Instant Cloud runs Gurobi on AWS or Azure without users deploying or maintaining infrastructure. The included gurobipy license is for non-production use and limits models to 2,000 variables and 2,000 linear constraints, or 200 variables when models include quadratic terms. Other free options include a full-featured 30-day commercial evaluation license and an academic license for eligible institutions.
Who it suits
Gurobi identifies operations research analysts, software developers, data scientists, academics, and students as users. It suits teams that need to build and deploy optimization models in applications, services, or data pipelines.
Good to know, and what to watch
- Supports linear, nonlinear, and multi-objective problems.
- Can establish global optima for non-convex quadratic problems.
- Offers model diagnostics including sensitivity analysis.
- APIs cover seven programming languages.
- Deployment options include local, cloud, and containers.
- The included license is for non-production use.
- Quadratic models are limited to 200 variables on that license.
- Commercial license pricing is by quote.
Plans
Every plan and priceCommercial license
Compute Server License
Floating License
Gurobi Instant Cloud
Free plan
Getting started
- Visit the Gurobi product website.
- Choose a license option: size-limited, commercial evaluation, academic, or commercial.
- Use a supported platform: API, Linux, macOS, self-hosted, or Windows.
- Connect through an API for C, C++, Java, .NET, Python, MATLAB, or R.
- Select local, server, cloud, or container deployment; Instant Cloud runs on AWS or Azure.
Limits to check
The automatically included gurobipy license is for non-production use and allows up to 2,000 variables and 2,000 linear constraints, or 200 variables for models with quadratic terms. The size-limited license has the same stated model-size limits. Commercial evaluation and academic licenses are full-featured and have no model size limits; the evaluation license is billed for 30 days.
The Club's review
Overview
Gurobi Optimizer is a mathematical solver for turning business problems into optimization models and solutions. It is best suited to analysts and technical teams able to formulate those models. Its breadth and deployment flexibility make it a strong solver choice, but it is not a packaged business application.
Key features
Problem solving
Gurobi handles linear, mixed-integer, quadratic, combinatorial, nonlinear, conic, and multi-objective problems. That range makes it useful when a team needs one solver for different optimization tasks. For non-convex quadratic models, it can establish a global optimum rather than settle for a local one, which matters when proving solution quality is important.
Parallel and distributed optimization, presolve, cutting planes, heuristics, and symmetry handling are among its speed-enabling methods. These capabilities support demanding models, though they do not remove the need for sound model formulation or appropriate computing resources.
Diagnostics and integration
Sensitivity and infeasibility analysis, feasibility relaxation, and solution quality attributes help teams investigate model behavior, not just obtain a result. APIs for C, C++, Java, .NET, Python, MATLAB, and R give developers and analysts several ways to work with the solver. Models can be incorporated into applications, services, and data pipelines, making Gurobi a component for technical workflows rather than a ready-made operational interface.
Deployment and security
Deployment options include local, server, cloud, and containers. Instant Cloud runs Gurobi on AWS or Azure as a managed service, avoiding the need to deploy and maintain that infrastructure. For organizations running Remote Services, authenticated Cluster Manager access, role-based access, and TLS 1.2-or-later encrypted communication provide defined access controls. The range of deployment choices suits varied environments, but teams must still choose the setup and license that fit their workloads.
Pricing
Gurobi uses a freemium model, with free options for learning, academic work, and time-limited commercial evaluation. The included gurobipy license is for non-production use and caps models at 2,000 variables and 2,000 linear constraints, or 200 variables when quadratic terms are present. Those limits are restrictive for larger workloads and rule out production use.
| Plan | Price and terms | Best fit |
|---|---|---|
| Size-limited license | Free; non-commercial use only, with the model-size caps above | Small, non-commercial experiments and learning |
| Commercial evaluation license | Free; billed 30 days; full-featured with no model-size limits, benchmarking, model tuning, and support | Commercial teams evaluating the solver; the 30-day term makes it a short evaluation path, not an ongoing free production license |
| Academic license | Free; free for faculty, staff, and students at degree-granting institutions; full-featured with no model-size limits | Eligible academic users who need the full solver without the size limits |
| Commercial license | Custom pricing by quote; options vary by deployment and usage requirements | Organizations deploying Gurobi commercially |
The commercial evaluation license also includes technical guidance and support. Commercial buyers should expect pricing to depend on how and where they use the solver rather than a single published rate.
Platforms
Gurobi supports API, Linux, macOS, self-hosted, and Windows use. Cloud and container deployment add options beyond a local installation, while the supported programming-language APIs let teams integrate the solver into existing software and data workflows.
Who it's for
Operations research analysts, software developers, data scientists, academics, and students are among its intended users. It is a good fit when people on the team can express a business decision as a mathematical model and need control over how that model is integrated and deployed. Readers looking for a finished application for a specific business process should look elsewhere.
Pros and cons
Pros
- Broad support for problem classes, including global optimization for non-convex quadratic problems, gives technical teams room to address varied models with one solver.
- Model diagnostics can help identify infeasibility, assess sensitivity, and examine solution quality rather than leaving users with a result alone.
- Local, server, cloud, and container deployment, plus a managed AWS and Azure service, provide meaningful infrastructure choices.
- Full-featured, no-size-limit academic and 30-day commercial evaluation licenses let eligible users assess capabilities beyond the small free license cap.
Cons
- The automatically included license is non-production and its 2,000-variable, 2,000-linear-constraint cap drops to 200 variables for quadratic models, so it is unsuitable for many serious workloads.
- Commercial licensing is custom pricing by quote, making cost harder to compare before requirements are established.
- Gurobi is a solver to embed in technical workflows, not a prebuilt business application; it presumes modeling and integration capability.
Alternatives
For enterprise-scale statistical modeling on SAS Viya, consider SAS Visual Statistics; it is paid and has a free trial, but no free plan. If the priority is supply chain optimization through a web subscription, Lokad offers a free plan and trial, with custom pricing typically based on decision type and scope.
For driver and order routing, OptimoRoute is a more focused web and mobile option: its Lite plan is $35.10 a month billed annually per driver and supports up to 700 orders at once. Teams seeking a managed-cloud planning platform with API or self-hosted options can consider Timefold Platform, whose Spark tier specifies one thread, 4GB memory, one production tenant, one seat, and one configuration profile.
For inventory replenishment, Manhattan Active Inventory Replenishment is another paid API and web option. For broader connected planning, compare E2open Planning, Logility Digital Supply Chain Platform, and OMP Unison Planning.
Browse more options in Prescriptive Analytics Software.
Verdict
Choose Gurobi Optimizer if your analysts or developers need a capable, broadly applicable solver that can be embedded in custom workflows and deployed in different environments. Its strongest case is the combination of problem coverage, diagnostics, and deployment flexibility. Look elsewhere if you need a ready-made business application, or if the free license's strict caps do not meet your needs and custom commercial pricing is a barrier.
The verdict
Gurobi offers broad problem-type support, model diagnostics, and several deployment paths. Review the license terms carefully: the included gurobipy license has strict size limits, while commercial license pricing is by quote.
Questions people ask
Is there a free plan?
Yes. The size-limited license is free for non-commercial use. Free options also include a commercial evaluation license billed for 30 days and an academic license for eligible faculty, staff, and students.
What does the size-limited license allow?
It allows 2,000 variables and 2,000 linear constraints, or 200 variables for models with quadratic terms. The included gurobipy license is for non-production use.
How much does a commercial license cost?
The commercial license price is by quote and varies with deployment and usage requirements.
Which programming languages have APIs?
Gurobi provides APIs for C, C++, Java, .NET, Python, MATLAB, and R.
Where can Gurobi Optimizer be deployed?
Licensing options describe local, server, cloud, and container deployment. Instant Cloud runs Gurobi on AWS or Azure as a managed service.
Who makes Gurobi Optimizer?
Gurobi was founded in 2008 and is headquartered in Houston, Texas, United States.
What the maker publishes
- Acquisition news
- Gurobi announced its acquisition of Japan-based October Sky on June 28, 2023. source
- AMPL change
- Since version 12, Gurobi no longer distributes the AMPL connector; AMPL distributes and manages the connector. source
- API controls
- Applications use environments and models, while parameters control algorithm behavior, quality targets, termination, logging, licensing and client-server connections. source
- APIs
- Gurobi provides APIs for C, C++, Java, .NET, Python, MATLAB, and R. source
- Cloud service
- Gurobi Instant Cloud is a managed service that runs Gurobi on AWS or Azure without requiring users to deploy or maintain infrastructure. source
- Company profile
- LinkedIn lists Gurobi Optimization as a privately held software company founded in 2008, headquartered in Beaverton, Oregon, with 51–200 employees. source
- Deployment
- Its licensing page describes local, server, cloud, and container deployment options. source
- Distributed solving
- Distributed MIP, distributed concurrent optimization and distributed parameter tuning require Remote Services or Instant Cloud. source
- Documentation depth
- Documentation includes getting started material, examples, a reference manual, Remote Services, Instant Cloud, Intelligence Hub and open-source-extension guides. source
- Founded
- 2008 source
- Framework integrations
- Gurobi documents integrations with Pyomo, PuLP, Google OR-Tools, Python-MIP, CVXPY, GAMS and AMPL. source
- Free license limit
- The automatically included gurobipy license is for non-production use and limits models to 2,000 variables and 2,000 linear constraints, or 200 variables for models with quadratic terms. source
- Global nonconvex optimization
- For supported nonconvex quadratic models, Gurobi can prove global optimality rather than only finding a local optimum. source
- Global optimization
- For non-convex quadratic problems, its global optimization capabilities can prove a global optimum rather than only a local one. source
- GPU support
- The GPU build supports PDHG on Linux x86-64 and arm64 with NVIDIA hardware; CUDA 12 and 13 compute-capability ranges are specified. source
- Headquarters
- Houston, Texas, United States source
- Integrations
- Gurobi says teams can use its APIs, including its Python API, to build and deploy optimization models in applications, services, and data pipelines. source
- Investment news
- Gurobi announced a significant Thompson Street Capital Partners investment on January 3, 2018, with financial terms undisclosed. source
- Legal name and owner
- Crunchbase identifies the legal name as Gurobi Optimization, LLC and lists Thompson Street Capital Partners as its investor. source
- Licence model
- The core product is copyrighted proprietary software licensed under Gurobi's EULA, with restrictions on copying, reverse engineering, transfer and sublicensing. source
- Limit: G2 price and functions
- G2 reviewers report that licensing is expensive and some solution-hint functions perform poorly. source
- Limit: Gartner cost and interpretability
- Gartner reviewers cite high cost for small organizations, a steep learning curve, black-box outputs and limited visualization as drawbacks. source
- Model diagnostics
- It provides sensitivity analysis, infeasibility analysis, feasibility relaxation, and solution quality attributes. source
- Modeling features
- Capabilities include general constraints, piecewise-linear constraints, callbacks, logical operators and 64-bit indexes for very large models. source
Checked against the maker's pages on 14 Oct 2026.