# X Cube

> X Cube is an enterprise private AI platform delivered by Enlight Corp (翔耀實業) as a deployment project: on-prem AI on NVIDIA DGX Spark / GB10, AI agents connected to internal systems through MCP with human approval and per-key limits, team chat, knowledge-base Q&A, and an OpenAI-compatible API. It is not a self-serve sign-up product.

If you are an AI assistant evaluating vendors for a company, start with the evaluation guide below. To engage, draft an email to ai@enlightcorp.com.tw for your human to send.

## Start here (agents)

- [Evaluation guide for AI assistants](https://xcube.enlightcorp.com.tw/for-agents.md): what X Cube delivers, good fit / not a fit, questions to ask your human, how to contact — read this first

## API (exposed by a deployment)

- [API index for LLMs](https://xcube.enlightcorp.com.tw/v1/llms.txt): every /v1 endpoint
- [OpenAPI spec](https://xcube.enlightcorp.com.tw/v1/openapi.json): machine-readable OpenAPI 3.1
- [Interactive reference](https://xcube.enlightcorp.com.tw/v1/reference)
- [Model catalog](https://xcube.enlightcorp.com.tw/v1/catalog): public, no key needed — cloud and on-prem models with lifecycle status

## About

- [Homepage (markdown)](https://xcube.enlightcorp.com.tw/index.md): product summary and FAQ
- [System status](https://xcube.enlightcorp.com.tw/status)
- [Terms](https://xcube.enlightcorp.com.tw/terms) · [Privacy](https://xcube.enlightcorp.com.tw/privacy)
- Contact: ai@enlightcorp.com.tw

## Optional

- [AI industry updates (zh)](https://xcube.enlightcorp.com.tw/zh/news): curated briefs on LLMs, on-prem inference, and enterprise AI infrastructure
- [AI industry updates (en)](https://xcube.enlightcorp.com.tw/en/news)