API Overview

The noris “LLM as a Service” API is fully OpenAI-compatible. The base URL for all endpoints is:

https://ai.noris.de/v1

Available Endpoints

Endpoint	Method	Description
/v1/models	GET	List available models
/v1/chat/completions	POST	Chat completions (incl. streaming, tool calling, vision)
/v1/completions	POST	Text completions (legacy)
/v1/responses	POST	OpenAI Responses API
/v1/embeddings	POST	Text embeddings
/v1/rerank	POST	Reranking (query/documents)
/v1/images/generations	POST	Image generation (if available)

Listing Models

A simple GET request returns all currently available models:

curl https://ai.noris.de/v1/models \
  -H "Authorization: Bearer YOUR_API_KEY"

A successful response contains the list of all currently available models:

{
  "object": "list",
  "data": [
    { "id": "vllm/release/gpt-oss-120b", "object": "model", "owned_by": "openai" },
    { "id": "vllm/release/gemma-4-31b-it", "object": "model", "owned_by": "google" },
    { "id": "vllm/release/harrier-oss-v1-0.6b", "object": "model", "owned_by": "microsoft" },
    { "id": "vllm/release/bge-reranker-v2-m3", "object": "model", "owned_by": "baai" }
  ]
}

The full and current model overview including tiers and specifications is available under Models.

Rate Limits

All endpoints are subject to rate limits (RPM and TPM) to ensure fair resource allocation. Limits scale with the customer’s commitment level. See Rate Limits for details.