Soft robotic manipulators (SRM) are biologically inspired from animal appendages such as elephant trunks and octopuses’ arms. In contrast to traditional rigid robotic manipulators, SRMs are made of flexible material and generate motion through structural deformation. Thus, they more easily adapt to unstructured environments. Teleoperated SRMs can be used in spaces harmful or impractical to humans (i.e., nuclear radiation or minimally invasive surgical sites). Limited research is available on the factors that affect human performance during the teleoperation of SRMs. We present two Human-Robot Interfaces (HRI) and conduct user studies on an SRM in object handling tasks and assess the performance of teleoperators using spatial input devices and augmented reality-based input devices. The user interaction is quantified for two types of controllers (Direct and Indirect) in an immersive pick and place operation. A System Usability Scale (SUS) questionnaire is administered to assess the usability of each HRI. Results suggest that the users perform more effectively, make fewer errors using the Indirect Control HRI and participants rated the Indirect Control HRI as more usable regardless of the hardware device.