JigsawStack is the AI SDK for every tech stack.

The JigsawStack AI SDK is a Typescript/Javascript library that allows you to interact with powerful AI services to build AI-powered applications in any framework like Next.js, React, Vue, Svelte, and more that supports JS.

• 🧩 Powerful AI services all in one SDK
• 🧑‍💻 Native JS/TS SDK
• ▲ Supports Next.js, React, Vue, Svelte, Node.js, Bun, Deno, and more
• ⌨️ Fully typed parameters and responses
• 📡 Built in Webhook support for long-running tasks
• 📦 Built in file system for easy file uploads
• 🍃 Lightweight and blazing fast, built ont native fetch with isomorphic-fetch for maximum compatibility

Learn more

To learn more about all available JigsawStack AI services, view the Documentation or Website.

All APIs

Learn more of about each category in the API reference

Installation

You will need Node.js 18+ on your local development machine.

npm install jigsawstack
yarn add jigsawstack

Setup

First, get your API key from the JigsawStack Dashboard

Then, initialize the SDK:

import { JigsawStack } from "jigsawstack";

const jigsaw = JigsawStack({ apiKey: "your-api-key" });

Usage

AI Scraping Example:

const resp = await jigsaw.web.ai_scrape({
    url: "https://www.amazon.com/Cadbury-Mini-Caramel-Eggs-Bulk/dp/B0CWM99G5W",
    element_prompts: ["prices"],
});

Text to Speech Example:

const resp = await jigsaw.audio.text_to_speech({
    text: "Hello, how are you doing?",
});

Upload a file to use across any API example:

const file = /*reference to file blob or buffer*/

await jigsaw.store.upload(file,{
    filename: "receipt.png",
});

// Now you can use the file_key in any API call
const ocrResp = await jigsaw.vision.vocr({
    file_store_key: "receipt.png",
});

const objectDetectionResp = await jigsaw.vision.object_detection({
    file_store_key: "receipt.png",
});

Community

Join JigsawStack community on Discord to connect with other developers, share ideas, and get help with the SDK.

Related Projects

• Docs
• Python SDK

Contributing

JigsawStack AI SDK is open-source and welcomes contributions. Please open an issue or submit a pull request with your changes. Make sure to be as descriptive as possible with your submissions, include examples if relevant.