We present panco2, an open-source Python library designed to extract galaxy cluster pressure profiles from maps of the thermal Sunyaev-Zeldovich effect. The extraction is based on forward modeling of the total observed signal, allowing to take into account usual features of millimeter observations, such as beam smearing, data processing filtering, and point source contamination. panco2 offers a large flexibility in the inputs that can be handled and in the analysis options, enabling refined analyses and studies of systematic effects. We detail the functionalities of the code, the algorithm used to infer pressure profile measurements, and the typical data products. We present examples of running sequences, and the validation on simulated inputs. The code is available on GitHub at https://github.com/fkeruzore/panco2, and comes with an extensive technical documentation to complement this paper at https://panco2.readthedocs.io.