🔍 Deep Dive: Jina AI - Chroma Docs

Chroma provides a convenient wrapper around JinaAI’s embedding API. This embedding function runs remotely on JinaAI’s servers, and requires an API key. You can get an API key by signing up for an account at JinaAI.

from chromadb.utils.embedding_functions import JinaEmbeddingFunction
jinaai_ef = JinaEmbeddingFunction(
    api_key="YOUR_API_KEY",
    model_name="jina-embeddings-v2-base-en",
)
jinaai_ef(input=["This is my first text to embed", "This is my second document"])

// npm install @chroma-core/jina

import { JinaEmbeddingFunction } from '@chroma-core/jina';

const embedder = new JinaEmbeddingFunction({
    jinaai_api_key: 'jina_****',
    model_name: 'jina-embeddings-v2-base-en',
});

// use directly
const embeddings = embedder.generate(['document1', 'document2']);

// pass documents to query for .add and .query
const collection = await client.createCollection({name: "name", embeddingFunction: embedder})
const collectionGet = await client.getCollection({name:"name", embeddingFunction: embedder})

You can pass in an optional model_name argument, which lets you choose which Jina model to use. By default, Chroma uses jina-embedding-v2-base-en.

Late Chunking Example

jina-embeddings-v3 supports Late Chunking, a technique to leverage the model’s long-context capabilities for generating contextual chunk embeddings. Include late_chunking=True in your request to enable contextual chunked representation. When set to true, Jina AI API will concatenate all sentences in the input field and feed them as a single string to the model. Internally, the model embeds this long concatenated string and then performs late chunking, returning a list of embeddings that matches the size of the input list.

from chromadb.utils.embedding_functions import JinaEmbeddingFunction
jinaai_ef = JinaEmbeddingFunction(
    api_key="YOUR_API_KEY",
    model_name="jina-embeddings-v3",
    late_chunking=True,
    task="text-matching",
)

collection = client.create_collection(name="late_chunking", embedding_function=jinaai_ef)

documents = [
    'Berlin is the capital and largest city of Germany.',
    'The city has a rich history dating back centuries.',
    'It was founded in the 13th century and has been a significant cultural and political center throughout European history.',
]

ids = [str(i+1) for i in range(len(documents))]

collection.add(ids=ids, documents=documents)

results = normal_collection.query(
    query_texts=["What is Berlin's population?", "When was Berlin founded?"],
    n_results=1,
)

print(results)

Task parameter

jina-embeddings-v3 has been trained with 5 task-specific adapters for different embedding uses. Include task in your request to optimize your downstream application:

retrieval.query: Used to encode user queries or questions in retrieval tasks.
retrieval.passage: Used to encode large documents in retrieval tasks at indexing time.
classification: Used to encode text for text classification tasks.
text-matching: Used to encode text for similarity matching, such as measuring similarity between two sentences.
separation: Used for clustering or reranking tasks.

Dense Embedding Models

Sparse Embedding Models

Frameworks

Jina AI

Late Chunking Example

Task parameter

​Late Chunking Example

​Task parameter

Late Chunking Example

Task parameter