News
Models
Products
keyboard_arrow_down
DeepSearch
Search, read and reason until best answer found.
Reader
Convert any URL to Markdown for better grounding LLMs.
Embeddings
World-class multimodal multilingual embeddings.
Reranker
World-class reranker for maximizing search relevancy.
More
keyboard_arrow_down
Classifier
Zero-shot and few-shot classification for image and text.
Segmenter
Cut long text into chunks and do tokenization.

API Docs
Auto codegen for your copilot IDE or LLM
open_in_new


Company
keyboard_arrow_down
About us
Contact sales
Intern program
Join us
open_in_new
Download logo
open_in_new
Terms & Conditions


Log in
login
Integrating Embeddings in RAG
Leveraging Jina Embeddings Superior Performance
Get Involved
Tech blog
December 06, 2023

Dify.AI integrates Jina Embeddings for RAG

Dify.AI, a leading open-source platform specialized in creating generative AI applications, is now leveraging Jina Embeddings v2!
Collaborative graphic of Jina x Dify in white text on a blue and purple wavy background
Scott Martens
Scott Martens • 3 minutes read

Online LLM application development platform Dify.AI has integrated the Jina Embeddings v2 API in its innovative AI toolkit for instant access when building and hosting LLM applications. All you need to do is add your Jina Embeddings API key via their intuitive web interface to get the full power of Jina AI’s industry-leading embedding models in your RAG (retrieval-augmented generation) applications.

Dify.AI x Jina AI:Dify now Integrates Jina Embedding Model - Dify Blog
The next-gen development platform - Easily build and operate generative AI applications. Create Assistants API and GPTs based on any LLMs.
Dify Blog
Embedding API
Top-performing, 8192-token length, $100 for 1.25B tokens, seamless OpenAI alternative, free trial

tagIntegrating Embeddings in RAG

Current AI architectures today have no direct way to integrate outside information sources. The model itself encodes information from its training data with varying levels of accuracy, and it is impractical to retrain the model every time there is new, potentially useful data that could be incorporated into it.

For example, I asked JinaChat a question about current events:

Jina AI chat application with a conversation about Henry Kissinger's status, a help disclaimer, and UI elements.

The only way to ensure that an LLM has the information needed to answer a factual question is to provide it in the prompt:

Chatbot screen with a message about Henry Kissinger's death on November 29, 2023, at age 100, citing the Washington Post.

Naturally, an LLM that only answers questions correctly if you include the answer in your question isn’t very useful. This has led to a body of techniques called Retrieval-Augmented Generation or RAG. RAG is a framework installed around an LLM that searches external information sources for materials that might contain the information needed to answer a user’s request and then presents them, with the user’s prompt, to the LLM.

Flowchart demonstrating a large language model process using "When did Henry Kissinger die?" as a user prompt example

This strategy has the added benefit that LLMs hallucinate much less when they are expected to handle a given text rather than recall things they might partially remember from training.

tagLeveraging Jina Embeddings Superior Performance

Dify.AI has integrated Jina Embeddings v2 to enhance retrieval quality for RAG prompting. Jina AI’s models provide state-of-the-art accuracy in RAG applications, and with an input window of 8,192 tokens, they can support much larger and more complex questions than most competing models at a much lower price.

You can now use Jina Embeddings in your LLM projects via Dify.AI’s intuitive application builder, as shown in the video below or in the post on Dify.AI's blog:

0:00
/0:38

tagGet Involved

Check out Dify.AI’s LLM application builder and hosting service for yourself. You can get a free tester token from the Jina AI website to use Jina Embeddings to try it out.

Dify.AI · The Innovation Engine for Generative AI Applications
The next-gen development platform - Easily build and operate generative AI applications. Create Assistants API and GPTs based on any LLMs.
Embedding API
Top-performing, 8192-token length, $100 for 1.25B tokens, seamless OpenAI alternative, free trial

For more information about Jina AI’s offerings, check out the Jina AI website or join our community on Discord.

Join the Jina AI Discord Server!
Check out the Jina AI community on Discord - hang out with 3873 other members and enjoy free voice and text chat.
Discord
Categories:
Tech blog
rss_feed

Read more
May 28, 2025 • 4 minutes read
Correlations: Vibe-Testing Embeddings in GUI
Jina AI
May 25, 2025 • 8 minutes read
Fair Scoring for Multimodal Documents with jina-reranker-m0
Nan Wang
Alex C-G
Stacked glowing green ovals on a background transitioning from black to green, with the top oval having an unusual, split sha
May 07, 2025 • 9 minutes read
Model Soup’s Recipe for Embeddings
Bo Wang
Scott Martens
Still life drawing of a purple bowl filled with apples and oranges on a white table. The scene features rich colors against a
Offices
location_on
Sunnyvale, CA
710 Lakeway Dr, Ste 200, Sunnyvale, CA 94085, USA
location_on
Berlin, Germany (HQ)
Prinzessinnenstraße 19-20, 10969 Berlin, Germany
location_on
Beijing, China
Level 5, Building 6, No.48 Haidian West St. Beijing, China
location_on
Shenzhen, China
402 Floor 4, Fu'an Technology Building, Shenzhen, China
Search Foundation
DeepSearch
Reader
Embeddings
Reranker
Classifier
Segmenter
API Documentation
Get Jina API key
Rate Limit
API Status
Company
About us
Contact sales
Newsroom
Intern program
Join us
open_in_new
Download logo
open_in_new
Terms
Security
Terms & Conditions
Privacy
Manage Cookies
email
Jina AI © 2020-2025.