An embedding maps a piece of content (a sentence, a paragraph, an image) to a fixed-length vector — typically 768, 1536, or 3072 dimensions. Items with similar meaning end up close in vector space, even if they share no words. This enables search by intent rather than keyword: query 'cheap places to fly to in summer' can match 'budget European vacations in July.' Embeddings power semantic search, recommendation engines, deduplication, clustering, RAG retrieval, and similarity-based classification. Modern embedding models (OpenAI's text-embedding-3, Voyage, Cohere's embed-v3) cost a fraction of LLM inference and run sub-100ms. Vector databases (Pinecone, Weaviate, Postgres pgvector) store and query embeddings at scale.
قاموس
ما هو Embedding؟
A vector representation of text, image, or audio — a list of numbers that captures the semantic meaning, enabling 'find similar' searches.
مصطلحات ذات صلة
Vector Database
A database optimized for storing and searching embeddings — finding the nearest vectors to a query vector in milliseconds, even across billions of records.
RAG (Retrieval-Augmented Generation)
A technique that lets an LLM answer questions about information it wasn't trained on — by retrieving relevant documents at query time and stuffing them into the prompt.
العودة إلى قاموس AI