How scenarios account for short-term nuclear power flexibility? Usual communication limited to Power and Energy or Capacity Factors. Need of details on cycling, speed, rated Pmin. First remark is that generally scenarios do not communicate on the way nuclear reactors adapt to vRES; they just show the installed capacities and CF (or the power generated). We find relevant indicator of the nuclear flexibility such as the number of cycles performed by NPP, their type and the speed of ramping, and the min rated power as well. We show that if reactors run faster, they better integrate RES, they substitute for gas and reduce system cost and power prices. But for the nuclear operator, this could come with a loss while evicting gas power form the market. A non-obvious result of particular interest is for the min rated power, as we obtain little sensitivity on the system by constraining the reactor to run down to 20% of the nominal power. This is because in 2050 (the year of simulation) we’ll have a small capacity of nuclear (25% compared to 75% currently). Similar to the point 2, Baseload is more benefical to the private investor, as it makes increases the power price. A larger share of the fleet that is flexible makes prices decrease, thus income will decrease. Final contribution is on the debate, how to diversify the energy mix and reduce the nuclear share, by phasing out reactors or by reducing the power generated with the existing reactors? We find that it makes more sense to reason in flows, as we still need a large capacity in 2050, in particular over seasons with low solar power, like winter and nights. Besides, this would limit the economic and social cost in terms of employment, regional dependency of industrial sites, and geographical power mix equilibria.