Quick start

Protect OpenAI spend with virtual keys and hard budget limits.

Quick Configuration Example

Base URL:: https://your-domain.com/api/v1/openai
API Key:: kg_live_xxxx
Model:: gpt-4o-mini

Replace the API key with your virtual key from the dashboard. Use a model supported by KeyGood and permitted by your virtual key's model allowlist.

Make your first API request

Copy the command below and replace the virtual key.

curl https://your-domain.com/api/v1/openai/chat/completions \
  -H "Authorization: Bearer kg_live_xxx" \
  -H "Content-Type: application/json" \
  -d '{
    "model": "gpt-4o-mini",
    "messages": [
      {
        "role": "user",
        "content": "Hello from KeyGood"
      }
    ]
  }'

The endpoint is OpenAI-compatible. Your virtual key is used for authentication and budget enforcement.

Environment / client config

OPENAI_BASE_URL=https://your-domain.com/api/v1/openai
OPENAI_API_KEY=kg_live_xxxx

Sign up
Create an account and sign in to the dashboard.
Bind your OpenAI API key
Settings → save your master key. It is encrypted (AES-256-GCM) and never shown again.
Create a virtual key
Set daily/monthly budgets, optional model allowlist and expiration. Copy the key once — only a hash is stored.
Point your client at KeyGood
Use the Base URL above and your kg_live_... key. Same paths as OpenAI, e.g. /chat/completions. Streaming is supported.
Hard Stop
When the budget is exhausted, the API returns 402 with budget_exceeded — the provider is not called.