Overview
Your Novion API at a glance.
Active keys–
Requests today–
Tokens · 7 days–
Daily limit per key–
Quickstart
Models
| ID | Name | Good for |
|---|
API keys
Keys are shown once when you create them. Store them somewhere safe — Novion keeps only a hash.
Copy your new key now. You won't be able to see it again.
| Name | Key | Created | Last used | Today | Status |
|---|
No keys yet. Create one above.
Playground
Test any Novion model. Playground calls count toward usage but don't use a key.
0.7
Send a message to start.
Usage
Last 7 days by model, and your 50 most recent calls.
By model · 7 days
| Model | Requests | Input tokens | Output tokens | Avg latency | Errors |
|---|
Recent calls
| Time | Model | Source | Tokens in/out | Latency | Status |
|---|
Docs
Base URL . Send your key as Authorization: Bearer KEY (or x-api-key: KEY).
OpenAI-compatible · POST /v1/chat/completions
Anthropic-compatible · POST /v1/messages
List models · GET /v1/models
Limits & notes
- Each key gets requests per day (resets at midnight UTC). Over the limit returns HTTP 429.
- max_tokens is capped at 4096. Streaming isn't supported yet — set stream to false or leave it out.
- If a model's engine is busy or over its limit, Novion quietly falls back to a smaller engine. The response's novion.engine field shows which one answered.
- Moonshot gives scenarios, not certainties. Nothing it says is financial advice.