LP
Linear programs
Continuous decisions with a linear objective and linear row bounds. HiGHS provides revised simplex, interior-point with crossover, and PDLP algorithms.
Loading runnable examples…
The open-source HiGHS optimization solver, packaged for Node.js and modern browsers. Build and solve linear, mixed-integer, and convex quadratic models from JavaScript and TypeScript.
ES module// Load the WebAssembly runtime import loadHighs from "highs"; const highs = await loadHighs(); // Solve a model in CPLEX LP format const result = highs.solve(lpText); console.log(result.ObjectiveValue);
What it solves
HiGHS is designed for models whose objective and constraints are linear, plus continuous convex quadratic objectives. Choose the class that matches which variables must be discrete and whether the objective contains quadratic terms.
LP
Continuous decisions with a linear objective and linear row bounds. HiGHS provides revised simplex, interior-point with crossover, and PDLP algorithms.
MILP
Linear models with integer, binary, or semi-continuous columns. HiGHS uses branch-and-cut with presolve, heuristics, and gap-based stopping.
QP
Continuous columns, linear constraints, and a convex quadratic objective supplied as a sparse Hessian. Non-convex QP and MIQP are outside HiGHS’ scope.
For a closer look
solve(lpText) · persistent createModel() · raw C interfaceLoading runnable examples…