GPTPort

Safely manage and share your OpenAI / Azure API keys.

Features

• OpenAI compatiable API - serve as a OpenAI proxy, thus protect your API keys from being leaked when using third-party Apps.
• Running fully on Edge - Supa FAST!
• Support Azure API - support import Azure API as LLM service provider.
• Auto expire - generate API keys that automatic expire after specific time.
• Multi user support - you can control who can sign up your service, both emails control and domain control are supported.

Environment Variables

To run this project, you will need to add the following environment variables to your .env file

Auth

GPTPort uses nextauth as authentication solution.

• NEXTAUTH_SECRET - a random secret for nextauth, you can generate one by running openssl rand -base64 48.
• NEXTAUTH_URL - you can omit this if deployed on Vercel.
• ADMIN_EMAIL - admin email.
• ALLOWLIST - only emails among this list can login, use * for public user, or domain name to allow all emails under specific domains.

Auth providers

You need at least one auth provider to get start, visit https://next-auth.js.org/providers/ for more information.

• GOOGLE_CLIENT_ID
• GOOGLE_CLIENT_SECRET

Storage

GPTPort uses @upstash/redis for storage, because it can run on the edge runtime and also offers a pretty generous free quota which should be more that enough for personal uses.

Visit https://upstash.com/ to get one, or deploy one with https://fly.io/docs/reference/redis/

• UPSTASH_REDIS_REST_TOKEN
• UPSTASH_REDIS_REST_URL

Authors

• @hopsken