Context. We study two coronal mass ejections (CMEs) observed between April 1 to 2, 2019 by both the inner Wide-Field Imager for Parker Solar Probe (WISPR-I) onboard the Parker Solar Probe (PSP) spacecraft (located between about 46 and 38 solar radii during this period) and the inner heliospheric imager (HI-1) onboard the Solar Terrestrial Relations Observatory Ahead (STEREO-A) spacecraft, orbiting the Sun at about 0.96 au. This is the first study of CME observations from two viewpoints in similar directions but at considerably different solar distances. Aims. Our objective is to derive CME kinematics from WISPR-I observations and to compare them with results from HI-1. This allows us to understand how the PSP observations affect the CME kinematics, especially due to its proximity to the Sun. Methods. We estimated the CME positions, speeds, accelerations, propagation directions, and longitudinal deflections using imaging observations from two spacecrafts and a set of analytical expressions that consider the CME as a point structure and take the rapid change in spacecraft position into account. We derived the kinematics using each viewpoint independently and both viewpoints as a constraint. Results. We found that both CMEs are slow (<400 km s−1), propagating eastward of the Sun-Earth line (westward of PSP and STEREO-A). The second CME seems to accelerate between ~0.1 and ~0.2 au and deflect westward with an angular speed consistent with the solar rotation speed. We found some discrepancies in the CME solar distance (up to 0.05 au, particularly for CME #1), latitude (up to ~10°), and longitude (up to 24°) when comparing results from different fit cases (different observations or set of free parameters). Conclusions. Discrepancies in longitude are likely due to the feature that is tracked visually, rather than instrumental biases or fit assumptions. For similar reasons, the CME #1 solar distance, as derived from WISPR-I observations, is larger than the HI-1 result, regardless of the fit parameters considered. Error estimates for CME kinematics do not show any clear trend associated with the observing instrument. The source region location and the lack of any clear in situ counterparts (both at near-Earth and at PSP) support our estimate of the propagation direction for both events.