Objective: This study aimed to explore the molecular response mechanisms of differential blood metabolites before and after 8 weeks of threshold and polarized training models using metabolomics technology combined with changes in athletic performance. Methods: Twenty-four male rowers aged 14–16 were randomly divided into a THR group and a POL group (12 participants each). The THR group followed a threshold training model (72%, 24%, and 4% of training time in low-, moderate-, and high-intensity zones, respectively), while the POL group followed a polarized training model (78%, 8%, and 14% training-intensity distribution). Both groups underwent an 8-week training program. Aerobic endurance changes were assessed using a 2 km maximal rowing performance test, and untargeted metabolome analysis was conducted to examine blood metabolomic changes before and after the different training interventions. Aerobic endurance changes were assessed through a 2 km maximal rowing test. Non-targeted metabolomics analysis was employed to evaluate changes in blood metabolome profiles before and after the different training interventions. Results: After 8 weeks of training, both the THR and POL groups exhibited significant improvements in 2 km maximal rowing performance (p < 0.05), with no significant differences between the groups. The THR and POL groups had 46 shared differential metabolites before and after the intervention, primarily enriched in sphingolipid metabolism, glutathione metabolism, and glycine, serine, and threonine metabolism pathways. Nine unique differential metabolites were identified in the THR group, mainly enriched in pyruvate metabolism, glycine, serine, and threonine metabolism, glutathione metabolism, and sphingolipid metabolism. A total of 14 unique differential metabolites were identified in the POL group, predominantly enriched in sphingolipid metabolism, glycine, serine, and threonine metabolism, aminoacyl-tRNA biosynthesis, and glutathione metabolism. Conclusions: The 8-week THR and POL training models demonstrated similar effects on enhancing aerobic performance in adolescent male rowers, indicating that both training modalities share similar blood metabolic mechanisms for improving aerobic endurance. Furthermore, both the THR group and the POL group exhibited numerous shared metabolites and some differential metabolites, suggesting that the two endurance training models share common pathways but also have distinct aspects in enhancing aerobic endurance.