Macroeconomic models inspired by the Cowles Commission's approach have displayed remarkable longevity in government economic agencies. They have weathered the criticisms directed against them in the literature. And thanks to their practical usefulness, they have survived the emergence and rise of other macroeconomic analytical tools. By striking a balance between their consistency with theory and the quality of their data fit, these generalist tools help to shed light on a wide range of issues that arise in an operating environment. Moreover, the modellers have incorporated the advances in time-series econometrics and now use the models in combination with other analytical tools. Developed by INSEE and the Directorate-General of the Treasury, Mésange is just such a model. It comprises some forty behaviour equations estimated econometrically, for a total of around 500 equations and three modelled sectors: manufacturing, non-manufacturing, and non-market. It exists in two versions, reflecting the two different methods for calculating volume data in the quarterly national accounts: volumes at constant prices and volumes at chain-linked prices. The latter correspond to the data published by INSEE since 2007. We describe and illustrate two types of applications. The first consists in analyzing the effect of shocks on the economy in the short, medium, and long terms. For this purpose, the model's long-term properties must be well known and controlled. This requirement leads users to prefer the version with volumes at constant prices. For the second application, which focuses on reviewing economic forecasts and analyzing the recent past, it is essential to fit the model to the data of the published accounts. Preference here goes to the version with volumes at chain-linked prices.