Purpose. Although the overall number of couples has decreased, Vietnamese couples seeking infertility treatment have increased in number. CAPA-IVM is not only a drug-free IVF treatment with a positive outcome, but also makes the entire IVF patient journey less complicated and causes less physical and emotional strain on the women by fewer injections causing no side-effects, and is also economical [1]. This study aimed to propose a machine-learning model using morpho-kinetics to predict CAPA-IVM embryo implantation and increase the rate of success for patients with a less invasive nature as PGT. Methods. This retrospective cohort study was conducted at IVFMD from March 2021 to March 2022 with 72 couples, aged ≤ 40 years old, and embryos cultured under a time-lapse system until the blastocyst stage. The model was built using Machine learning algorithms, including Keras, XGBoost, Logistic Regression, Random Forest, Decision Tree, and Support Vector Machine. Results. The best AUC score was 0.79 in the XGboost and Gradient boosting models. The t5_min, t4_min, t8_min, and tB_min were the four most important variables in the training model. Conclusions. Contingent on these predictions, the importance of morpho-kinetic selection of embryos, as it showed an increased number of pregnancies, a higher birth rate, and a lower loss of early pregnancies, which was confirmed by recent meta-analysis in IVF [2], has been reported firstly in CAPA IVM. Until now, the XGboost and Gradient boosting models seem efficient in predicting the quality of the good CAPA- IVM blastocyst embryo.