langchain-pixeltable 0.1.0

LangChain VectorStore integration for Pixeltable multimodal data infrastructure.

langchain-pixeltable

LangChain VectorStore integration backed by Pixeltable -- multimodal data infrastructure with built-in embedding indexes, incremental computation, and 25+ AI provider integrations.

Installation

pip install langchain-pixeltable

Quick Start

from langchain_pixeltable import PixeltableVectorStore
from langchain_openai import OpenAIEmbeddings

# Create a vector store and add documents
vs = PixeltableVectorStore.from_texts(
    texts=["Pixeltable handles multimodal data", "LangChain is great for chains"],
    embedding=OpenAIEmbeddings(),
    table_name="mydir.docs",
)

# Similarity search
results = vs.similarity_search("multimodal data management", k=3)
for doc in results:
    print(doc.page_content)

Connect to an Existing Pixeltable Table

vs = PixeltableVectorStore.from_existing_table(
    table_name="mydir.existing_docs",
    embedding=OpenAIEmbeddings(),
    text_column="content",
    embedding_column="content_embedding",
)
results = vs.similarity_search("search query")

Use as a LangChain Retriever

retriever = vs.as_retriever(search_kwargs={"k": 5})
docs = retriever.invoke("What is Pixeltable?")

Why Pixeltable as a Vector Backend?

• Persistent and versioned: Data survives restarts; every change is tracked
• Incremental: Only new/changed rows get re-embedded
• Multimodal native: Images, video, audio, and documents alongside text
• 25+ AI providers: Built-in functions for OpenAI, Anthropic, Gemini, and more
• No external services: Embedded PostgreSQL, no Docker required

Links

• Pixeltable Docs
• GitHub
• Discord