Title: One size does not fit all: Evidence for a range of mixing efficiencies in stellar evolution calculations

Abstract: Context: Internal chemical mixing in intermediate- and high-mass stars represents an immense uncertainty in stellar evolution models.In addition to extending the main-sequence lifetime, chemical mixing also appreciably increases the mass of the stellar core. Several studies have made attempts to calibrate the efficiency of different convective boundary mixing mechanisms, with sometimes seemingly conflicting results. Aims: We aim to demonstrate that stellar models regularly under-predict the masses of convective stellar cores. Methods: We gather convective core mass and fractional core hydrogen content inferences from numerous independent binary and asteroseismic studies, and compare them to stellar evolution models computed with the MESA stellar evolution code. Results: We demonstrate that core mass inferences from the literature are ubiquitously more massive than predicted by stellar evolution models without or with little convective boundary mixing. Conclusions: Independent of the form of internal mixing, stellar models require an efficient mixing mechanism that produces more massive cores throughout the main sequence to reproduce high-precision observations. This has implications for the post-main sequence evolution of all stars which have a well developed convective core on the main sequence.