Skip to main content
Kinetica is a database with integrated support for vector similarity search. It supports:
  • exact and approximate nearest neighbor search
  • L2 distance, inner product, and cosine distance
This notebook shows how to use the Kinetica vector store (Kinetica). This needs an instance of Kinetica which can easily be setup using the instructions given here - installation instruction. 
# Pip install necessary package
pip install -qU langchain-kinetica
We want to use OpenAIEmbeddings so we have to get the OpenAI API Key. 
import getpass
import os

from langchain_openai import OpenAIEmbeddings

if "OPENAI_API_KEY" not in os.environ:
    os.environ["OPENAI_API_KEY"] = getpass.getpass("OpenAI API Key:")

embeddings = OpenAIEmbeddings(model="text-embedding-3-large")
You must set the database connection in the following environment variables. If you are using a virtual environment you can set them in the .env file of the project:
  • KINETICA_URL: Database connection URL (e.g. http://localhost:9191)
  • KINETICA_USER: Database user
  • KINETICA_PASSWD: Secure password.
# Kinetica needs the connection to the database.
# Set these environment variables:
from gpudb import GPUdb

from langchain_kinetica import KineticaSettings, KineticaVectorstore

kdbc = GPUdb.get_connection()
k_config = KineticaSettings(kdbc=kdbc)
k_config

2026-02-02 21:28:34.745 INFO     [GPUdb] Connected to Kinetica! (host=http://localhost:19191 api=7.2.3.3 server=7.2.3.5)

KineticaSettings(kdbc=<gpudb.gpudb.GPUdb object at 0x1170ae270>, database='langchain', table='langchain_kinetica_embeddings', metric='l2')

from uuid import uuid4

from langchain_core.documents import Document

document_1 = Document(
    page_content="I had chocolate chip pancakes and scrambled eggs for"
    " breakfast this morning.",
    metadata={"source": "tweet"},
)

document_2 = Document(
    page_content="The weather forecast for tomorrow is cloudy and overcast"
    ", with a high of 62 degrees.",
    metadata={"source": "news"},
)

document_3 = Document(
    page_content="Building an exciting new project with LangChain - come check it out!",
    metadata={"source": "tweet"},
)

document_4 = Document(
    page_content="Robbers broke into the city bank and stole $1 million in cash.",
    metadata={"source": "news"},
)

document_5 = Document(
    page_content="Wow! That was an amazing movie. I can't wait to see it again.",
    metadata={"source": "tweet"},
)

document_6 = Document(
    page_content="Is the new iPhone worth the price? Read this review to find out.",
    metadata={"source": "website"},
)

document_7 = Document(
    page_content="The top 10 soccer players in the world right now.",
    metadata={"source": "website"},
)

document_8 = Document(
    page_content="LangGraph is the best framework for building stateful"
    ", agentic applications!",
    metadata={"source": "tweet"},
)

document_9 = Document(
    page_content="The stock market is down 500 points today due to"
    " fears of a recession.",
    metadata={"source": "news"},
)

document_10 = Document(
    page_content="I have a bad feeling I am going to get deleted :(",
    metadata={"source": "tweet"},
)

documents = [
    document_1,
    document_2,
    document_3,
    document_4,
    document_5,
    document_6,
    document_7,
    document_8,
    document_9,
    document_10,
]
uuids = [str(uuid4()) for _ in range(len(documents))]
uuids

['ddad79f1-141d-44f6-8f50-72e5c0f1ee16',
    '10819fa9-794b-4fde-934a-aabd453781c8',
    '3ce641d5-8c6b-4dcb-90fe-a3c19b3132ff',
    '9db5c865-389f-481c-aea2-440b8437e22c',
    '74dd4d80-a371-4c41-8254-7981d375274d',
    '74d7571e-f8c5-4001-9979-e99996ec2ce5',
    '3a3eb718-f2b9-4186-8c2e-34a1e18ebb3b',
    '59a88b08-f8c6-4cf5-b485-9485a4a8ffd0',
    'd84ad1c8-ec01-4d13-b61a-ef4b08abb485',
    'c9ab8f4f-e566-465f-a85d-ee05780714ea']

​
Similarity search with euclidean distance (Default)

The Kinetica Module will try to create a table with the name of the collection. So, make sure that the collection name is unique and the user has the permission to create a table. 
COLLECTION_NAME = "langchain_example"

vectorstore = KineticaVectorstore(
    config=k_config,
    embedding_function=embeddings,
    collection_name=COLLECTION_NAME,
    pre_delete_collection=True,
)

vectorstore.add_documents(documents=documents, ids=uuids)

print()
print("Similarity Search")
results = vectorstore.similarity_search(
    "LangChain provides abstractions to make working with LLMs easy",
    k=2,
    filter={"source": "tweet"},
)
for res in results:
    print(f"* {res.page_content} [{res.metadata}]")

print()
print("Similarity search with score")
results = vectorstore.similarity_search_with_score(
    "Will it be hot tomorrow?", k=1, emb_filter={"source": "news"}
)
for res, score in results:
    print(f"* [SIM={score:3f}] {res.page_content} [{res.metadata}]")

Similarity Search
* Building an exciting new project with LangChain - come check it out! [{'source': 'tweet'}]
* LangGraph is the best framework for building stateful, agentic applications! [{'source': 'tweet'}]

Similarity search with score
* [SIM=0.945353] The weather forecast for tomorrow is cloudy and overcast, with a high of 62 degrees. [{'source': 'news'}]

​
Working with vectorstore

​
Adding documents

Above, we created a vectorstore from scratch. However, often times we want to work with an existing vectorstore. In order to do that, we can initialize it directly. 
vectorstore = KineticaVectorstore(
    config=k_config,
    embedding_function=embeddings,
    collection_name=COLLECTION_NAME,
)

# We can add documents to the existing vectorstore.
vectorstore.add_documents([Document(page_content="foo")])

docs_with_score = vectorstore.similarity_search_with_score("foo")

print(f"First result: {docs_with_score[0]}")
print(f"Second result: {docs_with_score[1]}")

First result: (Document(metadata={}, page_content='foo'), 0.0014664357295259833)
Second result: (Document(metadata={'source': 'tweet'}, page_content='Building an exciting new project with LangChain - come check it out!'), 1.260981559753418)

​
Overriding a vectorstore

If you have an existing collection, you override it by doing from_documents and setting pre_delete_collection = True 
vectorstore = KineticaVectorstore.from_documents(
    documents=documents,
    embedding=embeddings,
    collection_name=COLLECTION_NAME,
    config=k_config,
    pre_delete_collection=True,
)

docs_with_score = vectorstore.similarity_search_with_score("foo")
docs_with_score[0]

(Document(metadata={'source': 'tweet'}, page_content='Building an exciting new project with LangChain - come check it out!'),
    1.2609236240386963)

​
Using a VectorStore as a retriever

from langchain_core.vectorstores.base import VectorStoreRetriever

retriever: VectorStoreRetriever = vectorstore.as_retriever()
retriever

VectorStoreRetriever(tags=['KineticaVectorstore', 'OpenAIEmbeddings'], vectorstore=<langchain_kinetica.vectorstores.KineticaVectorstore object at 0x1139cfaa0>, search_kwargs={})
Edit this page on GitHub or file an issue.
Connect these docs to Claude, VSCode, and more via MCP for real-time answers.