CVXPY Solvers

Scen-Opt delegates all optimization to CVXPY, which provides a unified interface to a wide range of open-source and commercial solvers. This page lists the solvers bundled with the Scen-Opt server and the full set of solvers supported by CVXPY.

Installed Solvers

The following solvers are installed on the Scen-Opt server and available out of the box. Any of these can be passed via the solver dropdown in the web interface or the solver argument in the Python API.

Solver

LP

QP

SDP

Notes

CLARABEL

X

X

X

Default solver for most problem types. Open-source, pure Rust.

SCS

X

X

X

Splitting conic solver. Handles all cone types including SDP.

MOSEK

X

X

X

Commercial solver (free academic license). High performance and reliability for LP, QP, and SDP.

ECOS

X

X

X

Lightweight interior-point solver.

CVXOPT

X

X

X

Python-native interior-point solver.

OSQP

X

X

Operator-splitting QP solver. LP and QP only.

DAQP

X

X

Dual active-set solver for LP and QP.

PIQP

X

X

Proximal interior-point QP solver.

HiGHS

X

X

High-performance LP/QP solver. Open-source.

SciPy

X

X

Uses SciPy’s linprog / minimize backends.

Tip

If no solver is specified, CVXPY automatically selects the most specialized solver for the problem type. For SDP problems, CLARABEL or SCS will be used. For QP problems, OSQP is typically preferred.

Additional Solvers

Beyond the solvers installed on the server, CVXPY supports many additional solvers that can be installed separately. The full compatibility matrix is shown below.

Solver

LP

QP

SDP

CBC

X

X

COPT

X

X

X

CPLEX

X

X

X

GLOP

X

GLPK

X

GLPK_MI

X

GUROBI

X

X

X

PDLP

X

PROXQP

X

X

QPALM

X

X

SCIP

X

X

X

SDPA

X

X

X

XPRESS

X

X

X

Note

Only LP, QP, and SDP columns are shown — these are the problem types supported by Scen-Opt. Solvers already listed in the Installed Solvers table above are omitted here.

For installation instructions and solver-specific options, see the CVXPY solver documentation.

References

[DB2016]

S. Diamond and S. Boyd, “CVXPY: A Python-embedded modeling language for convex optimization,” Journal of Machine Learning Research, vol. 17, no. 83, pp. 1–5, 2016.

[BV2004]

S. Boyd and L. Vandenberghe, Convex Optimization, Cambridge University Press, 2004.