Integration of Vision Transformer models with texture analysis presents a novel dual-stage approach for histopathology diagnosis. The proposed approach is mainly focused on discriminating between Meningioma (MEN) and Solitary Fibrous Tumor (SFT), two tumors known for their similar morphological characteristics. This approach leverages the inherent power of ViT models to capture global and local features from whole-slide images (WSIs), complementing this capability by integrating texture analysis techniques aimed at improving classification accuracy. Initially, ViT models are applied across three levels of WSI magnification, using rotationally and multi-scaled tiles to handle diverse scales and orientations inherent in histopathological imagery. ViT’s attention mechanisms capture intricate details and spatial correlations within WSIs, offering a comprehensive view of histological structures. Concurrently, texture analysis methods including 3D-CLBP, 3D-GLCM, and 3D-GLRLM are used to extract the inherent patterns of the WSIs, such as homogeneity/inhomogeneity, morphology, and connectivity alongside the three RGB channels to capture the influence of color features. The scores obtained at the output of both stages are then fused and passed to a deep neural network, enabling a more reliable diagnosis. The experimental results show an accuracy of 93.42%, sensitivity of 92.15%, specificity of 94.73%, precision of 94.74%, balanced accuracy of 93.44%, and F1 score of 93.42%. These results elucidate the potential of the proposed approach in enhancing histopathology diagnostics.