Abstract: China has the largest apple yield in the world, and harvesting is one of the most complex and least mechanized processes. Apple picking robots, equipped with autonomous operation capabilities, will contribute to the advancement of the apple industry. The picking manipulators is a key component of the picking robot, required a small size, light weight, large operational range, and high modularity. However, current apple picking manipulators often feature complex structures and limited modularity, making them unsuitable for multi-arm picking operations. To overcome these limitations, it is necessary to develop an apple picking manipulators with a larger range of motion, high modularity, and lightweight structure. Optimizing the configuration parameters of these manipulators is essential for achieving efficient, stable, and flexible operation. Existing researches primarily focus on optimizing parameters through space constraints rather than motion performance. In response, a modular configuration apple picking manipulator was designed and an optimization design method was proposed to optimize the parameters of the manipulator. Firstly, based on the distribution of fruits in orchards and operational requirements, a specialized apple picking manipulator was designed. The manipulator consisted of three translational joints and three rotational joints. The three translational joints control motion along the x, y, and z axes, while the three rotational joints control rotation along the roll, pitch, and yaw axes. The horizontal joint and telescoping joint achieved different types of motion in the xoy plane by controlling joint drive motor 1 and 2 with translational motion. Next, multiple indices were constructed to evaluate the manipulator’s operational accessibility, structural compactness, velocity smoothness, and load smoothness. The multiple indices were combined into a single objective function. The analytic hierarchy process was employed to determine the weights of each index, and linear weighting was used to generate a comprehensive objective function. Then, an optimization algorithm based on an improved hippopotamus optimization algorithm (IHOA) was proposed. This hybrid algorithm employed hippopotamus optimization algorithm (HO) for global search in the initial stage, utilized particle swarm optimization (PSO) to accelerate convergence through collaboration and learning within the population, and incorporated simulated annealing (SA) to introduce random perturbations. Finally, simulation and field experiments were performed to validate the operational reachability, structural compactness, velocity stability, and load stability of the apple picking manipulator. The simulation results showed that the link lengths of pitch joint, horizontal joint, telescoping joint, end-effector revolute joint were 122.02, 138.00, 101.45 and 103.12 mm, respectively. The link offsets of telescoping joint, end-effector revolute joint, end-effector prismatic joint, twisting joint were 855.00, 166.67, 189.95 and 126.63 mm, respectively. The installation heights of the lower picking manipulator and the upper picking manipulator were 1344.59 and 2460.00 mm, respectively. The experimental results showed that operational accessibility index F₁, structural compactness index F₂, global velocity fluctuation performance index F₃, and global load fluctuation performance index F₄ were 97.05%, 2882.74 mm, 0.20 m/s and 0.15 N·m, respectively. Field experiments showed that during picking apples at the boundary points, the maximum absolute torque increments for the pitch joint, joint motor 1 and 2 with translational motion, and end rotation joint were 0.51, 0.87, 0.80, and 0.79 N·m, respectively. The maximum absolute velocity increments were 0.03, 0.17, 0.17, and 0.01 m/s, respectively. Within an operational range of 890.25 to 1035.47 mm from the tree trunk, the manipulator demonstrated full accessibility to boundary positions in the apple harvesting area. This study will contribute to the mechanization of apple harvesting, improving efficiency, reducing labor costs and providing valuable insights for the design of modular picking manipulators.