Skip to main content

AI glossary Models & Architecture

What is a vector in AI?

← All glossary terms
In short

A list of numbers that represents a piece of data, such as a sentence or an image, so that similar content ends up with similar numbers.

Explained

In AI, a vector is an ordered list of numbers that describes something in a form a machine can compare. An embedding model turns text, images or other data into vectors, usually with hundreds or thousands of dimensions, arranged so that items with similar meaning sit close together. Measuring the distance between vectors then becomes a way of measuring how similar two pieces of content are. This is what makes semantic search possible, and it is the core of retrieval in RAG systems, where the question and the documents are compared as vectors.

How content becomes a vector

An embedding model reads the input and outputs a fixed-length vector. The individual numbers have no meaning on their own; what matters is the position of the vector relative to others. Two sentences that say the same thing in different words get vectors that are close, while unrelated sentences get vectors that are far apart.

The same idea works for images, audio and code, and multimodal models can place different types of content in the same vector space.

What vectors are used for

  • Semantic search: find documents by meaning, not only by matching keywords.
  • RAG: retrieve the passages closest to a question and pass them to a language model.
  • Recommendations: suggest items whose vectors are close to what a user has liked.
  • Clustering and deduplication: group similar items or detect near-duplicates.
  • Anomaly detection: flag items whose vectors are far from the usual pattern.

Vector, embedding and vector database

The terms are related but not the same. A vector is the mathematical object, a list of numbers. An embedding is a vector produced by a model to represent meaning. A vector database stores large numbers of these vectors and finds the closest ones to a query quickly, using approximate nearest-neighbour indexes.

In everyday AI conversation, "vector" and "embedding" are often used interchangeably.

Frequently asked questions

What is the difference between a vector and an embedding?

A vector is any ordered list of numbers. An embedding is a specific kind of vector, produced by a model so that its position reflects the meaning of the content. All embeddings are vectors, but not all vectors are embeddings.

How is similarity between vectors measured?

The most common measures are cosine similarity, which compares the angle between two vectors, and Euclidean distance, which compares how far apart they are. Many systems also use the dot product. The choice usually follows what the embedding model was trained for.

What are vector dimensions?

Dimensions are the number of values in each vector. More dimensions can capture finer distinctions in meaning but take more storage and computation. Embedding models offer different sizes, and some let you choose.

What is vector search?

Vector search finds the stored vectors closest to a query vector, which means finding the content most similar in meaning. It is the retrieval method behind most semantic search and RAG systems, often combined with keyword search.

Do I need a vector database to use vectors?

Not for small collections, which can be searched in memory or with a vector extension in an existing database. A dedicated vector database becomes useful with large volumes, frequent updates, filters by metadata or strict latency needs.

Where it shows up at Thinkia

Further reading

Sources

Updated: 2026-10-21

Want to see how this applies to your organisation?

Talk to an AI Expert