Dictionary
Embedding Dimensions
The vector length of an embedding model's output.
1 min readupdated 2026-07-04
/ quick answer
Common dimensions: 384 (small), 768 (base), 1536 (OpenAI text-embedding-3-small), 3072 (large). Higher dimensions capture more nuance at the cost of storage and query speed. The vector length of an embedding model's output.
The vector length of an embedding model's output. Common dimensions: 384 (small), 768 (base), 1536 (OpenAI text-embedding-3-small), 3072 (large). Higher dimensions capture more nuance at the cost of storage and query speed. In practice: OpenAI's text-embedding-3-large returns 3072-dim vectors but supports truncation to 1536 or 512. This dictionary node is part of the Onexial knowledge graph and links to related concepts, workflows and tools below.
Definition
Common dimensions: 384 (small), 768 (base), 1536 (OpenAI text-embedding-3-small), 3072 (large). Higher dimensions capture more nuance at the cost of storage and query speed.
Example
OpenAI's text-embedding-3-large returns 3072-dim vectors but supports truncation to 1536 or 512.
Related Tool Stacks
/ frequently asked
What is Embedding Dimensions?
Common dimensions: 384 (small), 768 (base), 1536 (OpenAI text-embedding-3-small), 3072 (large). Higher dimensions capture more nuance at the cost of storage and query speed.
What is an example of Embedding Dimensions?
OpenAI's text-embedding-3-large returns 3072-dim vectors but supports truncation to 1536 or 512.
Why does Embedding Dimensions matter for AI and automation?
The vector length of an embedding model's output. It connects to the workflows, prompts and tool stacks linked on this page, so you can move from definition to execution without leaving Onexial.
↳ connected nodes
Tool Stack↳ linked
RAG Starter Stack
Minimum viable stack to ship a production RAG chatbot.
Dictionary↳ linked
Embedding
A numerical vector representation of text, image or audio that captures meaning for similarity search.
Dictionary↳ linked
Retrieval
Selecting the most relevant chunks for a query before generation.
Dictionary↳ linked
RAG (Retrieval-Augmented Generation)
Inject external knowledge into an LLM at query time.
Dictionary↳ linked
Agentic RAG
RAG where an agent decides what to retrieve, when, and from which source — instead of a single static query.
Dictionary↳ linked
Reranker
A second-stage model that reorders retrieved chunks for precision.