QUERY one endpoint, whichever model we choose

An OpenAI-compatible endpoint that is ours

Register, verify your email, mint a key, change one line in your app. What answers behind it is a config switch, not a rewrite.

Use it

from openai import OpenAI

client = OpenAI(
    base_url="https://query.ai-ministries.com/v1",
    api_key="qk-your-key-here",
)

client.chat.completions.create(
    model="dolphin3:8b",
    messages=[{"role": "user", "content": "hello"}],
)

curl works the same way:

curl https://query.ai-ministries.com/v1/chat/completions \
  -H "Authorization: Bearer qk-your-key-here" \
  -H "Content-Type: application/json" \
  -d '{"model":"dolphin3:8b","messages":[{"role":"user","content":"hello"}]}'

What you get

GET /v1/models and POST /v1/chat/completions, streaming included. Per-key rate limits and a usage log you can see. GET /healthz reports the live state with no key needed.