Supadata

This page covers how to use the Supadata ecosystem within LangChain. It allows you to load data from YouTube, TikTok, and other platforms.

Installation and setup

To use Supadata, you need to install the standalone integration package.

npm install @supadata/langchain-js

Document Loader

The Supadata integration allows you to load content from various video platforms as LangChain documents. It supports fetching transcripts and metadata.

Usage

import { SupadataLoader } from "@supadata/langchain-js";

const loader = new SupadataLoader({
  apiKey: "YOUR_SUPADATA_API_KEY", // or process.env.SUPADATA_API_KEY
});

const docs = await loader.load({
  url: "https://www.youtube.com/watch?v=dQw4w9WgXcQ",
  operation: "transcript",
  lang: "en",
  text: true,
  mode: "auto",
});

console.log(docs[0].pageContent);
console.log(docs[0].metadata);

Metadata Example

You can also fetch video structured metadata.

import { SupadataLoader } from "@supadata/langchain-js";

const loader = new SupadataLoader({
  apiKey: "YOUR_SUPADATA_API_KEY",
});

const docs = await loader.load({
  url: "https://www.youtube.com/watch?v=dQw4w9WgXcQ",
  operation: "metadata",
});

console.log(docs[0].pageContent);

Resources

NPM Package: @supadata/langchain-js
GitHub Repository: supadata-ai/langchain-js

For detailed usage examples and API documentation, please refer to the official GitHub repository.