Indium gallium arsenide (In1−xGaxAs) is an ideal material choice for short wave infrared (SWIR) imaging due to its low dark current and excellent collection efficiency. By increasing the indium composition from 53% to 83%, it is possible to decrease the energy gap from 0.74 eV to 0.47 eV and consequently increase the cutoff wavelength from 1.7 μm to 2.63 μm for extended short wavelength (ESWIR) sensing. In this work, we apply our well-established numerical modeling methodology to the ESWIR InGaAs system to determine the intrinsic performance of pixel detectors. Furthermore, we investigate the effects of different buffer/cap materials. To accomplish this, we have developed composition-dependent models for In1−xGaxAs, In1−xAlxAs, and InAs1−y Py. Using a Green’s function formalism, we calculate the intrinsic recombination coefficients (Auger, radiative) to model the diffusion-limited behavior of the absorbing layer under ideal conditions. Our simulations indicate that, for a given total thickness of the buffer and absorbing layer, structures utilizing a linearly graded small-gap InGaAs buffer will produce two orders of magnitude more dark current than those with a wide gap, such as InAlAs or InAsP. Furthermore, when compared with experimental results for ESWIR photodiodes and arrays, we estimate that there is still a 1.5x magnitude of reduction in dark current before reaching diffusion-limited behavior.