Pre-trained deep learning models can integrate artistic styles into photographic content, which is a promising neural style transfer technique. This study used pre-trained CNNs and Vision Transformers (ViTs) to mix creative styles with photographic pictures to stylize everyday shots. This method collects content image structural details and art image stylistic features by using the feature extraction capabilities of these models. After computing the content and style losses from several layers of pre-trained models, an iterative optimization procedure minimizes the overall loss and creates a visually appealing synthesis of content and style. Experimental results using impressionism and cubism show that the proposed method maintains crucial content qualities while incorporating unique creative elements such as color palettes and brush strokes. The results show that pre-trained models improve stylization and minimize computational needs compared with training models from scratch. This technology opens up new possibilities for creative picture modification in digital painting, photo editing, and augmented reality.