Objective To investigate the clinical value of magnetic resonance imaging (MRI) radiomics model in predicting chemotherapy resistance in breast cancer patients.Methods 240 breast cancer patients admitted from January 2021 to December 2024 were selected and divided into training set (160 cases) and verification set (80 cases) in chronological order. All patients received neoadjuvant chemotherapy regimens based on taxanes and anthracyclines. Chemotherapy resistance was evaluated based on residual cancer burden (RCB) index and tumor shrinkage rate after treatment completion. A total of 851 radiomics features were extracted from MRI images using the PyRadiomics platform, and 10 most predictive features were selected through LASSO regression. Clinical model, radiomics model, and combined model were constructed, with multi-level validation strategies employed to assess model performance. Results Among 240 patients, 58 developed chemotherapy resistance, with a resistance rate of 24.17% (25.00% in training cohort, 22.50% in validation cohort). The resistant group showed significantly higher positive expression rates of estrogen receptor (ER), progesterone receptor (PR), human epidermal growth factor receptor 2 (HER-2), and Ki-67 compared to the sensitive group (P<0.05). Multivariate analysis revealed that ER positivity, Ki-67≥20%, relative mean kurtosis (rMK), relative mean diffusion (rMD), mean apparent diffusion coefficient (ADC) value, and dose intensity completion rate were independent predictors of chemotherapy resistance. The combined model achieved AUCs of 0.891 (95%CI:0.842-0.940) and 0.879 (95%CI: 0.806-0.952) in training and validation cohorts respectively, with sensitivities of 87.50% and 88.89%, and specificities of 88.33% and 85.48%, significantly outperforming individual clinical and radiomics models (P<0.05). Bootstrap validation demonstrated good stability and consistency of the model. Conclusion The MRI radiomics-based combined predictive model can effectively predict chemotherapy resistance in breast cancer patients, providing objective and accurate imaging evidence for clinical individualized treatment decision-making