<i>Context</i>. The earliest phases of the formation of high-mass stars are not well known. It is unclear whether high-mass cores in monolithic collapse exist or not, and what the accretion process and origin of the material feeding the precursors of high-mass stars are. As outflows are natural consequences of the accretion process, they represent one of the few (indirect) tracers of accretion. Aims. We aim to search for individual outflows from high-mass cores in Cygnus X and to study the characteristics of the detected ejections. We compare these to what has been found for the low-mass protostars, to understand how ejection and accretion change and behave with final stellar mass.<br><i>Methods</i>. We used CO (2–1) PdBI observations towards six massive dense clumps, containing a total of 9 high-mass cores. We estimated the bolometric luminosities and masses of the 9 high-mass cores and measured the energetics of outflows. We compared our sample to low-mass objects studied in the literature and developed simple evolutionary models to reproduce the observables.<br> <i>Results</i>. We find that 8 out of 9 high-mass cores are driving clear individual outflows. They are therefore true equivalents of Class 0 protostars in the high-mass regime. The remaining core, CygX-N53 MM2, has only a tentative outflow detection. It could be one of the first examples of a true individual high-mass prestellar core. We also find that the momentum flux of high-mass objects has a linear relation to the reservoir of mass in the envelope, as a scale up of the relations previously found for low-mass protostars. This suggests a fundamental proportionality between accretion rates and envelope masses. The linear dependency implies that the timescale for accretion is similar for high- and low-mass stars.<br><i>Conclusions</i>. The existence of strong outflows driven by high-mass cores in Cygnus X clearly indicates that high-mass Class 0 protostars exist. The collapsing envelopes of these Class 0 objects have similar sizes and a similar fragmentation scale to the low-mass equivalents, and have enough mass to directly form high-mass stars from a monolithic collapse. If the pre-collapse evolution is quasistatic, the fragmentation scale is expected to limit the size of the initial mass reservoirs for all masses leading to higher densities at birth and therefore shorter free-fall times for higher mass stars. However, we find the collapse timescales to be similar for both lowand high-mass objects. This implies that in a quasi-static view, we would require significant turbulent/magnetic support to slow down the collapse of the more massive envelopes. But with this support still to be discovered, and based on independent indications of large dynamics in pre-collapse gas for high-mass star formation, we propose that such an identical collapse timescale implies that the initial densities, which should set the duration of the collapse, should be similar for all masses. Since the fragmentation scale is identical for all masses, a lower initial density requires that the mass that incorporates massive stars has to have been accreted from larger scalesthan those of low-mass stars and in a dynamical way.< Réduire