Zorveus Documentation

Send your first AI request through Zorveus using the standard OpenAI SDK and verify usage in your dashboard.

This guide walks you through setting up a startup organization in Zorveus, generating an inference key, and sending your first AI completion.

Test Key InserterInteractive

Paste your Zorveus key to dynamically update all code snippets below.

1. Prerequisites

Before writing code, complete these 3 steps in the Zorveus Dashboard:

Create a Startup Organization

Register an App

Navigate to Apps and create a new application (e.g. My AI Assistant).

Generate an Inference Key

Under your app's Inference Keys tab, click Create Key. Copy and save the generated key (zrv_...) immediately.

One-Time Secret Display

Inference keys are returned in plaintext only upon creation or rotation. Zorveus stores only cryptographic hashes and cannot retrieve the raw secret later.

2. Send Your First Request

Zorveus is 100% compatible with the official OpenAI SDKs across all major languages. You only need to set baseURL to https://api.zorveus.com/v1 and use your zrv_... key.

curl https://api.zorveus.com/v1/chat/completions \
  -H "Content-Type: application/json" \
  -H "Authorization: Bearer zrv_your_inference_key" \
  -d '{
    "model": "openai/gpt-4.1-mini",
    "messages": [
      {"role": "system", "content": "You are a helpful coding assistant."},
      {"role": "user", "content": "Write a Python function to check for prime numbers."}
    ],
    "temperature": 0.7
  }'

3. Verify Observable Results

After running the code:

Open your dashboard at app.zorveus.com.
Navigate to Usage & Analytics.
You will observe a new usage event recording:
- Request timestamp and status (200 OK)
- Prompt and completion token counts
- Provider cost and Zorveus wallet debit
- Attribution to your app and active inference key

Model permissions, spending caps, and routing rules are configured server-side inside Zorveus. The client SDK does not send billing caps or routing directives—the gateway enforces them automatically based on the resolved zrv_... key.

Next Steps

Product-User Metering

Attribute inference requests to individual customers using external_user_id.

Bring Your Own Key

Route requests through your own provider credentials to avoid Zorveus markup.

Was this page helpful?

Edit this page on GitHub

Startup Quickstart