Time | Content |
February 2026 | The Optimization Solver SDK is updated to V2.3.0: Migrated to new algorithm architecture for improved performance and scalability. Added MISOCP (Mixed-Integer Second-Order Cone Programming) solver module. Improved MPS/LP I/O performance for faster model loading and writing. Enhanced IPM (Interior Point Method) and PDHG algorithm performance. Added solution pool feature to PowerMILP for retrieving multiple solutions. Added support for cvxpy modeling language integration.
|
August 2025 | The Optimization Solver SDK is updated to V2.2.0: Added a Primal-Dual Method (PDM) algorithm that supports GPU acceleration. Added six parameters to control the PDM algorithm and three parameters to tune the Mixed-Integer Programming (MIP) heuristic algorithm. Added a new Mixed-Integer Linear Programming (MILP) module (Power Milp), which is enabled by default in most scenarios. Supports solving multi-objective optimization problems and reading or writing multi-objective MPS model files. Supports the feasibility relaxation (feasrelax) feature. Added related interfaces and examples.
|
May 2025 | The Optimization Solver SDK is updated to V2.1.2: Outputs an error message when a floating license verification fails (40x series error). Improved the user experience of installing the Python SDK from the installation package. Improved the performance of the data masking feature.
|
April 2025 | The Optimization Solver SDK is updated to V2.1.1: Improved the performance of the select interface in the Python SDK TupleDict class. Added more detailed banner information, including commit hash, build time, and platform. Supports function auto-completion for the Python SDK in Integrated Development Environments (IDEs) such as VSCode and PyCharm. Fixed known algorithm issues. Fixed an issue where floating licenses were not released correctly in some scenarios.
|
February 2025 | The Optimization Solver SDK is updated to V2.1.0: Added a Community Edition (CE) license. The license is no longer embedded in the pip package. Added a cluster solving authorization service that supports cluster deployment and multi-machine computation. Added support for Python 3.12 and 3.13. Dropped support for Python 3.6 and 3.7. Added the IPM/NumericFocus parameter. Modified the NumericFocus, EnableStochasticLP, and PostScaling parameters.
|
November 2024 | The Optimization Solver SDK is updated to V2.0: Added a MATLAB software development kit (SDK) with corresponding documentation and examples. Added modeling support for Quadratically Constrained Programming (QCP), Mixed-Integer Quadratic Programming (MIQP), and Mixed-Integer Quadratically Constrained Programming (MIQCP) problems. Added the MIP/EnableLazyConstr and NumericFocus parameters. Removed the MIP/NumericFocus parameter.
|
July 2024 | The Optimization Solver SDK is updated to V1.3: Added the MIP/NumericFocus parameter to control the level of attention paid to numerical precision for MIP problems. Added code examples for the Java and C# SDKs. Improved algorithm performance and fixed bugs.
|
May 2024 | The Optimization Solver SDK is updated to V1.2.1: Added the OutputFlag, LogToConsole, and LogFile parameters to control log output. Fixed a bug that caused an excessive number of iterations when solving some Quadratic Programming (QP) problems.
|
April 2024 | The paid version (subscription) of the solver is launched for the first time. It includes versions for both online and offline environments.Purchase>>. For more information, see Pricing and billing. The Optimization Solver SDK is updated to V1.2.0 (Download V1.2.0): Added the PostScaling parameter and improved MILP performance.
|
January 2024 | The Optimization Solver SDK is updated to V1.1.x (Download V1.1.1): Improved the performance of MILP and QP algorithms. Added five tunable parameters for MILP. Added the SolutionNumber parameter to retrieve multiple suboptimal solutions from the MIP solving process. Added a C# API with documentation and examples. Supports invoking MindOpt in Julia using the JuMP and AmplNLWriter.jl packages.
|
October 2023 | The Optimization Solver SDK is updated to V1.0.x (Download V1.0.1): Introduced a brand-new MindOpt SDK API and corresponding code examples. This version is not compatible with V0.x APIs. Added features for MILP, including warm start, SOS constraints, and Indicator constraints. Added a callback function feature. Added support for the .qps input file format. Provides a primal-dual feasible solution for linear programming problems.
|
October 2023 | New feature (Free Edition) published: MindOpt Copilot AI Engineer. An optimization technology consulting service robot based on a Large Language Model (LLM). It communicates about business problems through text chat and table files to automatically generate mathematical modeling formulas, write code, call software for solving, and interpret the results. |
August 2023 | The Optimization Solver SDK is updated to V0.25.x (Download V0.25.1): Improved the performance of the interior-point method. Improved MILP performance and added several configurable integer parameters. Added a MILP Infeasible Irreducible Subsystem (IIS) algorithm. Improved IIS stability, attribute retrieval, and error correction.
|
April 2023 | New feature (Free Edition) published: MindOpt Tuner Tuner. Based on one or a set of optimization problem data, it tunes the hyperparameters of the solver to achieve faster solving speeds and a better optimizer GAP. |
March 2023 | The Optimization Solver SDK is updated to V0.24.x (Download V0.24.1): Improved the performance of the MILP solver. This is a formal release version. All cloud authentication accounts are granted full access permissions. You no longer need to apply for a special license. Access is being rolled out automatically. If your access is not yet active, contact us. Improved the performance of the concurrent optimization method. Redesigned the Cholesky factorization in the IPM. Added a command-line feature for data masking. This allows for safer testing with external data. For more information about how to use this feature, see Data masking. Added installation packages for Linux ARM and Apple Silicon. Added the pip install mindoptpy installation method.
|
February 2023 | New feature (Free Edition) published: A new version of MindOpt APL Modeling Language is online. MindOpt Algebraic Programming Language (MAPL) is an algebraic modeling language that expresses mathematical models uniformly and efficiently. You can use it to easily switch between and call various solvers. |
November 2022 | The Optimization Solver SDK is updated to V0.23.x (Download V0.23.1): Added APIs related to Semidefinite Programming (SDP). Added support for the dat-s input file format. Fixed some bugs in MILP solving and improved performance. Added an LP installation package for Apple Silicon.
|
August 2022 | The Optimization Solver SDK is updated to V0.20.x (Download V0.20.2): Added a branch-and-cut solver for solving Mixed-Integer Linear Programming (MILP) problems. A limited number of test licenses are available. To apply, send an email request. Optimized the installation package for a simpler installation process. Optimized license verification. A new license-free logic is available: Optimization problems with fewer than 300 variables and 300 constraints can be solved without a license file (mindopt.lic or fl_client.ini).
|
May 2022 | The Optimization Solver SDK is updated to V0.19.x (Download V0.19.0): Added support for solving convex Quadratic Programming (QP) problems in Nonlinear Programming (NLP). Improved the performance of the simplex and interior-point methods. Improved the features and stability of the AMPL interface and the concurrent method. Redesigned license management. Added APIs such as createEnv to support large-scale, high-concurrency, and low-latency calls.
|
November 2021 | The Optimization Solver SDK is updated to V0.15.x (Download V0.15.1): Added the network simplex method and the multi-threaded simplex method. Improved the stability of the simplex and interior-point methods. Updated the data reading module and modified several APIs. The new network flow method performs well in the Large Network-LP Benchmark evaluation. For more information, see the details.
|
September 2021 | The Optimization Solver product entered public preview. Optimization Solver SDK V0.14.0 (Download V0.14.0) was published to solve Linear Programming (LP) problems. For beginners, see Quick Start to get started. |