This paper presents three spherical vector-based optimization techniques, namely the spherical vector-based spider wasp optimizer (SSWO), the spherical vector-based secretary bird optimization algorithm (SSBOA), and the spherical vector-based improved spider wasp optimizer (SISWO), to properly plan UAV trajectories in 3D complicated environments with various threats. SISWO is based on combining some SBOA stages with SWO to benefit from their strengths in dealing with local optima and accelerating convergence speed. Six scenarios generated in Christmas Island, Australia, are used to assess the effectiveness of the proposed algorithms in optimizing four different objectives, including path optimality, threat cost, flight height, and smooth cost. In addition, they are compared to seven recent and well-established algorithms according to several performance metrics. According to the experimental results, both SISWO and SSBOA could outperform all other algorithms in most scenarios, demonstrating that they are more effective at precisely planning the UAV flight path in complex 3-D environments. Quantitatively, in terms of Friedman’s mean rank, SISWO could achieve an average rank of 2.04 for all scenarios, followed by SSBOA with 2.27.