« Back to Glossary Index

An embedding model is an AI model that converts text, images, or other data into a list of numbers — a vector — capturing the underlying meaning in a form a computer can compare mathematically. Text with similar meaning ends up positioned close together in that numerical space, even if the actual wording is completely different.

Embeddings let a machine understand that "puppy" and "young dog" are closely related concepts, even though the words share no letters at all — something that's simply impossible with old-fashioned keyword matching.

  • Semantic search — surfacing genuinely relevant results even when exact keywords don't match
  • AI chatbots — helping a chatbot pull the most relevant information from a knowledge base
  • Recommendation systems — finding genuinely similar products or content
  • Content-based knowledge bases — powering "AI trained on your own content" plugins and tools

Embedding models sit under the hood of AI plugins like AI Engine, which use them to build a searchable knowledge base directly from a site's own posts and pages — letting an AI chatbot answer visitor questions using your actual content, rather than generic outside information.

« Back to Index
Share This