ngff-zarr

---

A multi-language implementation of the Open Microscopy Environment (OME) Next Generation File Format (NGFF) Zarr specification.

Repository Structure

This repository contains multiple packages implementing NGFF-Zarr support:

• py/ - Python package (ngff-zarr) - A lean and kind NGFF-Zarr implementation
• ts/ - TypeScript/JavaScript package (@fideus-labs/ngff-zarr) - Universal OME-Zarr for Deno, Node.js, and browsers
• mcp/ - Model Context Protocol (MCP) server (ngff-zarr-mcp) for AI integration

Python Package (py/)

The main Python package provides:

✨ Features

• Minimal dependencies
• Work with arbitrary Zarr store types
• Lazy, parallel, and web ready -- no local filesystem required
• Process extremely large datasets
• Conversion of most bioimaging file formats
• Multiple downscaling methods
• Supports Python>=3.10
• Reads OME-Zarr v0.1 to v0.5 into simple Python data classes with Dask arrays
• Optional OME-Zarr data model validation during reading
• Writes OME-Zarr v0.4 to v0.5
• Sharded Zarr stores
• Optional writing via zarr-python 2, zarr-python 3, tensorstore or zarrita (TypeScript)
• Anatomical orientation metadata (RFC-4)
• OME-Zarr Zip (.ozx) file support for single-file OME-Zarr datasets (RFC-9)
• High Content Screening (HCS) support for plate and well data
• Model Context Protocol (MCP) server for AI agent integration

📖 Documentation

More information about command line usage, the Python API, library features, and how to contribute can be found in our documentation.

TypeScript/JavaScript Package (ts/)

The TypeScript package provides universal OME-Zarr support for modern JavaScript environments:

✨ Features

• 🦕 Deno-first with first-class TypeScript support
• 📦 Universal compatibility - Works in Deno, Node.js, and browsers
• 🔍 Type-safe with Zod schema validation
• 🗂️ OME-Zarr v0.4 and v0.5 support
• 🌐 Web ready - No filesystem dependencies, works with remote stores
• 🏗️ Mirrors Python API - Familiar interfaces for Python users
• 📚 Lazy loading - Efficient handling of large datasets

📦 Installation

# Deno
import * as ngffZarr from "jsr:@fideus-labs/ngff-zarr";

# Node.js / npm
npm install @fideus-labs/ngff-zarr

# Browser (CDN)
import * as ngffZarr from "https://esm.sh/@fideus-labs/ngff-zarr";

📖 Documentation

See the TypeScript documentation for detailed usage, API reference, and examples.

MCP Server (mcp/)

The Model Context Protocol server enables AI assistants like Claude to interact with NGFF-Zarr files. See the MCP documentation for setup and usage instructions.

See also

• ome-zarr-py
• multiscale-spatial-image
• itk-ioomezarrngff
• iohub
• pydantic-ome-ngff
• aicsimageio
• bfio
• zod-ome-ngff

License

ngff-zarr is distributed under the terms of the MIT license.