API
Inference API
Use one OpenAI-compatible request format across model deployments, public models and LLM Model Routing endpoints.
One request shape
Send chat completions, embeddings and supported OpenAI parameters to https://api.qdiv0.com/v1. The model field selects a deployment, a public model or a routing endpoint.
inference.py
1from openai import OpenAI
2
3client = OpenAI(
4 base_url="https://api.qdiv0.com/v1",
5 api_key="qdiv0_sk_...",
6)
7
8response = client.chat.completions.create(
9 model="your-serving-name",
10 messages=[{"role": "user", "content": "Hello"}],
11)
12
13print(response.choices[0].message.content)Choose the backend
Use a deployment serving name for capacity-based workloads, a catalog identifier for metered models, or an LLM Model Routing serving name when QDivZero should choose the destination.
Operational controls
Streaming, tool use and LLM Firewall are request-level capabilities. Enable them without introducing another gateway or changing SDKs.