OpenAI-compatible endpoints, first-class streaming, and pricing you can predict to the cent.
Point your existing OpenAI SDK at our base URL — most apps migrate by changing two lines.
# pip install openai from openai import OpenAI client = OpenAI(api_key="novo-...", base_url="https://api.novoriollc.business/v1") res = client.chat.completions.create( model="novo-4-pro", messages=[{"role": "user", "content": "Hello!"}], ) print(res.choices[0].message.content)
Cached input is discounted automatically. No minimum commitment.
| Model | Input (per 1M) | Output (per 1M) | Cache hit |
|---|---|---|---|
| Novo-4-Pro | €0.55 | €2.20 | €0.10 |
| Novo-4-Flash | €0.09 | €0.36 | €0.02 |
Prices in EUR, Sep 2026. Volume tiers available for >1B tokens/month.
Server-sent events with sub-second first-token latency on Novo-4-Flash.
Guaranteed valid JSON output with strict schema enforcement on both models.
Parallel tool calls with reliable argument parsing, tested on 40k+ real workflows.
Full-context retrieval quality verified with our published LongBench-v2 suite.
Free evaluation keys include 10M tokens. No credit card required.