Online viscosity information on processing lines can reflect the material flow resistance and offer valuable guidance for manufacturing across various industries. Considering the accuracy, devices, and processes involved in injection molding, characterizing the melt's flow state during material processing poses a significant challenge. To reduce investment in viscometers, avoid influencing the components' surface aesthetics due to the installation of sensors, and make the flow state detect online in mold, this study designs a rheometric mold with cylindrical runners for identifying the in situ viscosity of molten resin during injection molding. The detection conditions of injection speed and cavity pressure variations, the entrance effect, and the viscous dissipation for Polycarbonate are analyzed under various conditions. The in situ viscosity is identified and compared with the standard cross‐WLF model. The result shows that the melt velocity and cavity pressure variations during the filling process create a stable environment for in situ rheological characterization and the detected viscosity is related to the shear rate, melt temperature, and channel dimension in injection molding. The designed mold with cylindrical runners for determining the in situ thermal‐rheological behavior of polymer is distinguished successfully and exhibits prospects for the development of low‐cost, nondestructive, and inner‐mold measurement in manufacturing applications.