<p>The Black-winged Kite Algorithm (BKA) is a novel method inspired by the attacking and migration behaviors of black-winged kites, which offers excellent solutions for optimization problems. However, its strategy limitations can lead to premature or repetitive convergence, leading instability across multiple runs. To address these issues, this paper introduces the Multi-strategy Black-winged Kite Algorithm (MSBKA), which improves the performance of BKA by integrating multiple advanced strategies. First, an adaptive parameter adjustment mechanism balances exploration and exploitation effectively. In the attacking phase, the introduction of the Lévy flight strategy diversifies search patterns, while the dynamic opposition-based learning strategy leverages opposing information to improve attacking behavior. During the migration phase, a random Black-winged Kite swarm strategy steers migration and reduces randomness, which ensures more consistent outcomes. Additionally, a dynamic diversity elite archive maintains solution diversity and enables the fine-tuning of population positions. A novel boundary handling strategy that combines randomness with a reflection mechanism further minimizes information loss. Experimental evaluations on the CEC2017 and CEC2022 benchmark suites demonstrate that MSBKA outperforms BKA by achieving more stable and superior results, as well as surpassing other algorithms in overall performance. Ablation studies validate the contribution of each strategy, while excellent performance on five real-world optimization problems underscores the algorithm’s engineering applicability. Moreover, a real-world optimization task on an image enhancement model further verifies the strong practical optimization capability of MSBKA and its ability to achieve superior enhancement results.</p>