Z3

Z3 is a high-performance theorem prover and satisfiability modulo theories (SMT) solver developed by Microsoft Research. It is designed to reason about logical formulas and to support software verification, program analysis, and formal methods. Z3 implements a DPLL(T) architecture, combining Boolean reasoning with a collection of theory solvers that handle arithmetic, arrays, uninterpreted functions, and other theories. It supports the SMT-LIB 2 standard and provides a flexible tactic framework that enables composing solvers for different tasks and optimization.

The solver supports a broad range of theories, including linear real and integer arithmetic, bit-vectors, arrays, uninterpreted sorts and functions, algebraic data types, strings, sequences, and floating-point arithmetic. It offers model finding, quantifier handling via techniques such as MBQI, and various optimizations to scale to large verification conditions. Z3 can operate as a standalone solver or be embedded in applications.

Interfaces include C, C++, .NET, Python, and Java APIs, with the Python API (Z3Py) being widely used in research and education. Users can interact with Z3 through SMT-LIB scripts or through the APIs to build, assert, and solve formulas, then retrieve satisfiability results and constructed models. It is cross-platform, running on Windows, Linux, and macOS.

Z3 was created by Leonardo de Moura and Nikolaj Bjørner at Microsoft Research. It is open-source, released under the MIT License, and maintained as a community-friendly project on GitHub. Z3 is widely used in software verification, formal methods, and automated reasoning across academia and industry.