Demonstration of DEMO relevant fusion power (P fus) level and tritium self-sufficiency are two important goals of the China fusion engineering testing reactor (CFETR). In this work the integrated modeling including self-consistent core–pedestal coupling are used to design the hybrid scenario plasmas at flat-top phase for these goals. Such plasmas have been taken as the reference plasma for studying the compatibility of the hybrid scenario with CFETR engineering design in the past two years. The physics justification for the selection of plasma density, Z eff, safety factor profile, and in particular the choice of auxiliary heating and current drive is presented. According to a scan of plasma density and Z eff, the target of P fus ≈ 1 GW and finite ohmic flux consumption ∆Φohm (4 h) ⩽ 250 Vs can be met with Z eff = 1.9–2.2 and the density at the pedestal top set at 90% of the Greenwald limit. Turbulent transport analysis using the gyro-Landau-fluid model TGLF shows that the electromagnetic effects can enhance the energy confinement but reduce the particle confinement and thus P fus. A baseline hybrid scenario case matching the target in the concept design is built using a combination of neutral beams (NB) and electron cyclotron (EC) waves to flatten the safety factor profile in the deep core region (with the normalized plasma radius ρ ⩽ 0.4). Such profile can yield better particle and energy confinement than that with either higher magnetic shear in the deep core region or higher q value in outer core region (e.g., due to the addition of lower hybrid current drive). Switching a part of auxiliary heating from electron to ions, e.g., replacing a part of EC waves by waves in the ion cyclotron range of frequencies, reduces the particle confinement and thus P fus. Since high harmonic fast waves (HHFW) can drive current at the same location as ECCD with higher current drive efficiency than ECCD and yield more electron heating than NB, the case using HHFW to replace a part of EC waves and NB can yield higher P fus and lower ∆Φohm than the baseline case. A discussion is given on future simulations to explore the improvement in plasma performance and the broadening of the feasible design space.