HAHA is a novel approach for generating animatable human avatars from monocular input videos. The method combines Gaussian splatting and a textured mesh to achieve high-fidelity rendering. It learns to apply Gaussian splatting only where necessary, such as for hair and out-of-mesh clothing, reducing the number of Gaussians used and minimizing rendering artifacts. This allows for the animation of small body parts like fingers, which are traditionally challenging to handle. The method is evaluated on two open datasets: SnapshotPeople and X-Humans. On SnapshotPeople, HAHA achieves comparable reconstruction quality to state-of-the-art methods while using less than a third of Gaussians. On X-Humans, it outperforms previous methods in terms of both quantitative and qualitative metrics, particularly in handling novel poses and views. The main contributions of HAHA include the combination of Gaussians and textured mesh, an unsupervised method for reducing the number of Gaussians, and the ability to efficiently animate highly articulated parts without additional engineering.HAHA is a novel approach for generating animatable human avatars from monocular input videos. The method combines Gaussian splatting and a textured mesh to achieve high-fidelity rendering. It learns to apply Gaussian splatting only where necessary, such as for hair and out-of-mesh clothing, reducing the number of Gaussians used and minimizing rendering artifacts. This allows for the animation of small body parts like fingers, which are traditionally challenging to handle. The method is evaluated on two open datasets: SnapshotPeople and X-Humans. On SnapshotPeople, HAHA achieves comparable reconstruction quality to state-of-the-art methods while using less than a third of Gaussians. On X-Humans, it outperforms previous methods in terms of both quantitative and qualitative metrics, particularly in handling novel poses and views. The main contributions of HAHA include the combination of Gaussians and textured mesh, an unsupervised method for reducing the number of Gaussians, and the ability to efficiently animate highly articulated parts without additional engineering.