Developed a modular integration of TimM pre-trained models into the brain-score/vision repository, enabling seamless use of ConvNeXt variants and Vision Transformers for computer vision benchmarking. The work involved designing a scalable extension point by registering timm-backed models and implementing utilities for model loading, preprocessing, and configuration management. Leveraging Python and PyTorch, the integration allows researchers to efficiently incorporate a wide range of deep learning models into existing evaluation pipelines. This addition not only expands the benchmarking capabilities of brain-score/vision but also lays the groundwork for future model family integrations, supporting ongoing advancements in machine learning and computer vision research.