Aistor

Source: https://github.com/minio/aistor

Description

Aistor is the official MCP server for AIStor, serving as a reference implementation for MCP server capabilities. It is designed as a hyperscale object store for AI, allowing enterprises to consolidate all their data into a single, private cloud namespace. Aistor is architected with hyperscaler principles to deliver scalable performance, primarily targeting Kubernetes environments.

Features

• Hyperscale object storage system designed for AI workloads
• Enables consolidation of enterprise data into a single, private cloud namespace
• Runs natively on Kubernetes across multiple providers
• Operator-based deployment (install all or select individual operators)
• Designed for performance at scale, optimized for enterprise use cases
• Reference implementation for MCP (MinIO Control Plane) server capabilities
• Open-source codebase

Installation

• Deployable via kubectl for operator installation
• Supports installation with Helm charts
• Requires a license secret (trial license available via MinIO subscription learning platform)

Category

Reference Implementations & Examples

Tags

reference-implementation, open-source, mcp, examples

Pricing

• 60 days trial license available via registration on the MinIO subscription learning platform
• Further licensing details available at https://subnet.min.io/