What problem does it solve? Classifying images, searching image collections by text, or moderating visual content normally requires training custom models on labeled data. CLIP removes that requirement by matching images and text in a shared embedding space, enabling zero-shot vision tasks with plain natural language labels. ## Core Features & Use Cases - Zero-Shot Image Classification: Classify images into arbitrary text categories without any training data or fine-tuning. - Semantic Image Search: Index image embeddings and retrieve matches using natural language queries, with optional integration into vector databases like Chroma or FAISS. - Content Moderation: Score images against safety categories such as NSFW or violent content using softmax probabilities. - Use Case: Given a folder of product photos, encode them once with ViT-B/32, then let users search the catalog with queries like "a red sneaker on white background" and return ranked results. ## Quick Start Use the CLIP skill to classify the attached image against the labels dog, cat, bird, and car and show the confidence scores.
