Introducing text-embedding-3-smallOpenAIReleased January 25, 2024

text-embedding-3-small

Highly efficient embedding model featuring Matryoshka dimension shortening for ultra-low storage and compute costs.

embeddingsProprietary API$0.02 / 1M tok (1536 dims)Context: 8.191K (8,191 tokens)

Technical Specifications

Architecture Type
Dense Matryoshka Representation Transformer
Total Parameters
Compact Dense
Context Window
8.191K (8,191 tokens)
Max Output Tokens
1.536K (1,536 tokens)
Knowledge Cutoff
September 2021
Supported Modalities
text
License & Access
OpenAI Business Terms

Benchmark Evaluations

Mteb
62.3
Miracl
44

Deep Architectural Overview

text-embedding-3-small generates 1536-dimensional embeddings for vector search and RAG. Using Matryoshka Representation Learning, developers can shorten vectors to 512 dimensions without substantial loss of retrieval accuracy.

Strengths & Considerations

Core Strengths
  • Ultra-cheap $0.02 per 1M tokens
  • Matryoshka representation shortening down to 512 dims
  • Outperforms text-embedding-ada-002
Known Limitations
  • Complex multi-lingual search is better served by text-embedding-3-large

Token & API Pricing

Input Tokens (1M)$0.02
Output Tokens (1M)Contact
Pricing is verified directly against OpenAI's developer documentation and API rate sheets.

Similar & Alternative Models

Explore other frontier models from OpenAI and comparable reasoning engines.

Browse all models
OpenAI

OpenAI GPT-6 model for focused, high-volume workloads where cost efficiency is the priority.

1.05M ctxUSD 0.10 / 1M input tok; USD 0.50 / 1M output tok
OpenAI

OpenAI GPT-6 model built for complex coding and agentic workflows, balancing intelligence and cost.

1.05M ctxUSD 2 / 1M input tok; USD 10 / 1M output tok
OpenAI

OpenAI’s premier sixth-generation frontier model, built for the most demanding end-to-end software engineering and scientific research tasks.

1.05M ctx$10.00 / 1M tok (GPT-6 Frontier)