A power system dominated by renewable energy is one of the key measures for achieving carbon neutrality. Demand response (DR) is a promising flexible resource for alleviating the supply-demand matching of high-proportion renewable energy systems. With the application of modern technologies, the potential for residential DR is growing. Electricity price is the key to improving residential DR capacity. However, existing dynamic pricing programs may fail to motivate end-users to adjust demand based on fluctuations in wind and photovoltaic (PV) output. This study proposes a dynamic pricing model that combines the fluctuation characteristics of residential electricity demand and wind and PV output, and utilizes bi-level optimization to coordinately dispatch the flexible loads. A case study of smart residential community consisting of 200 households shows that dynamic pricing incentivizes residential consumers to shift flexible loads from morning and evening to noon or early morning, which effectively improves the degree of matching between wind and PV output and residential electricity demand. Moreover, bi-level optimization effectively alleviates the potential rebound peak caused by large-scale residential participation in DR and achieves a relatively flat net grid demand profile. Furthermore, the dynamic pricing can incentivize residential consumers to participate in DR by reducing electricity bills. Display omitted •Building a dynamic pricing model for characterizing renewable energy fluctuation.•Proposing a bi-level optimization method for residential demand response.•Demand response reduces the supply-demand imbalances of renewable energy.•Electricity bills for residential customers are saved.