Embedding Models/Jina Embeddings v3

Jina Embeddings v3 by Jina AI — 1024D

Multilingual embeddings with task-specific LoRA adapters. Optimized for text retrieval, semantic matching, and long documents up to 8K tokens.

At a glance

Modalities

Dimensions

1K

Max tokens

8K

Parameters

559M

Price / 1M tokens

$0.020

Type

Dense

Matryoshka dimensions

32641282565121024

Output types

Single VectorMulti Vector

Language support

🌍 Multilingual support (89+ languages)

Benchmarks

MTEB EN 54.33
MMTEB 58.58
LONGEMBED 55.66

Details

Release date 2024-09-18
License CC BY-NC 4.0
Model ID jina-embeddings-v3
Provider Jina AI

Tags

text-embeddingmultilinguallong-contextmatryoshkalora-adapters

What You Need to Know About Jina Embeddings v3

Complete Specifications for Jina Embeddings v3 by Jina AI

Get detailed specifications for Jina Embeddings v3, including output dimensionality of 1024 dimensions, maximum input token length, supported input modalities, and pricing per million tokens. This embedding model is designed for semantic search, text classification, clustering, and retrieval-augmented generation applications where understanding the relationship between texts is essential.

Pricing and Cost Efficiency for Jina Embeddings v3

Review the pricing structure for Jina Embeddings v3 and compare it against other embedding models from Jina AI and competitors. Understanding embedding model costs is essential when scaling vector search and RAG applications to millions of documents. We provide transparent pricing to help you budget effectively.

Use Cases and Applications for Jina Embeddings v3

Explore the ideal use cases for Jina Embeddings v3. Whether you are building a semantic search engine, recommendation system, document classification pipeline, or multilingual retrieval system, understanding this model's capabilities and dimensionality will help you choose the right embedding strategy for your project.