AI API Relay Station Step-by-Step Setup: From Personal Tool to Monetized Service
\
\
Project Description: This article is compiled from a user-submitted (hook tutorial) and is for learning and information reference only. Online information is complex and scammers are everywhere; please think independently and judge carefully based on your own situation.
\
\
Published: 2026-05-10
\
\
2026 latest AI API relay station setup tutorial: Based on the New-API open-source project, learn how to purchase a server from scratch, deploy Docker, configure channels and the admin panel, and turn a personal tool into a profitable API relay service.
\
\
Tutorial Core Content Summary
\
Core Information Overview
\
\
\
Project Essence
\
This is a technical tutorial on building an AI API relay station based on the open-source project New-API. The core process involves purchasing an overseas server, deploying Docker containers, configuring upstream AI channels (such as OpenAI), and then reselling API credits to ordinary users at a higher price to earn the margin (volume difference or rate difference). Essentially, it is a side-business model that uses technical tools for API proxy distribution.
\
\
\
\
Core Methods & Workflow
\
\
- Purchase an overseas server that supports trials (such as Japan/US/Hong Kong nodes; a 2-core 4GB configuration is recommended, with a one-click image preinstalled with the BT Panel)
\
- Log in to the BT Panel, search for and one-click install the New-API project (official image) in the Docker app store, and open the external access port (default 3000)
\
- Bind a custom domain through the BT Panel reverse proxy feature, with the backend pointing to local port 3000, to enable public access
\
- Log in to the New-API admin panel (default admin/admin123), go to Channel Management, and add an upstream API channel (fill in the key, API address, and select the model group)
\
- Configure it in System Settings > Group Model Pricing