Agricultural supply chains face mounting challenges in balancing economic efficiency with social and environmental sustainability, particularly under conditions of supply and price uncertainty. Jujube is a strategic horticultural product in Iran, but its current supply chain suffers from high waste, weak coordination, and lost value-adding opportunities. This paper presents a multi-objective mixed-integer linear programming model for designing a sustainable closed-loop jujube supply chain based on circular economy principles. The model simultaneously minimizes total cost, maximizes job creation, and improves responsiveness. Responsiveness is measured through the reliability of processing and distribution facilities, so the model selects more reliable nodes when it allocates flows. The environmental dimension is improved by converting all agricultural residues and processing waste into biochar and by imposing a strict CO₂ emission cap on all echelons of the network. Uncertainty in orchard capacity and purchase prices, and their inverse relationship, is modeled by a two-stage stochastic programming approach with optimistic, realistic, and pessimistic scenarios. The proposed framework is applied to a real case study in South Khorasan Province, which produces about 98% of Iran’s jujube. The network includes several types of processing facilities and multiple domestic market segments. The stochastic multi-objective model is solved by the AUGMECON2 method in order to generate the complete Pareto front and reveal the trade-offs among cost, employment, and responsiveness. Results show that the integrated design can significantly reduce total cost while increasing both job opportunities and environmental performance compared with current fragmented practices. The analysis of Pareto solutions highlights the high social and service-level cost of aggressive cost minimization and helps decision-makers select balanced policies. Sensitivity analyses on demand, shortage cost, transport cost, and the carbon cap, as well as on the presence of the biochar facility, confirm that the results are robust. These tests show that the model is practically useful for strategic decision support in agricultural supply chains under uncertainty.