This project presents an AI-based clothing classification system that identifies the type of garment from an input image using the Fashion-CLIP zero-shot model. Traditional wardrobe or outfit recommendation applications depend on manual tagging of clothing items, making the process slow, inconsistent, and user-dependent. To overcome these limitations, this system uses Fashion-CLIP, a vision–language model trained on large-scale fashion data, to automatically classify clothing without requiring any custom training. The model extracts visual features from the input image and compares them against text-based labels such as “T-shirt,” “jeans,” “dress,” and “shirt,” enabling instant zero-shot classification with high accuracy. The proposed system supports image uploads, processes them through the Fashion-CLIP pipeline, and outputs the most probable clothing type along with confidence scores. This approach eliminates the need for dataset preprocessing, model training, or manual labeling, making classification efficient and scalable. The system can be integrated with wardrobe recommendation platforms, e-commerce apps, or personal styling tools. Overall, the project demonstrates a simple yet powerful method for automatically identifying clothing types using advanced vision–language AI models, reducing human effort and enabling smart, real-time fashion understanding.