Quickstart

This page shows the shortest path from an API key to a working request.

Prerequisites

• A Xenodia account.
• A long-term Xenodia API key.
• A server environment that can keep secrets out of the browser.

Set your key:

export XENODIA_API_KEY="YOUR_LONG_TERM_KEY"

Send a chat request

curl -X POST "https://api.xenodia.xyz/v1/chat/completions" \
  -H "Content-Type: application/json" \
  -H "Authorization: Bearer $XENODIA_API_KEY" \
  -d '{
    "model": "openai/gpt-4o-mini",
    "messages": [
      { "role": "system", "content": "You are a precise test assistant." },
      { "role": "user", "content": "Reply with OK only." }
    ],
    "temperature": 0
  }'

Expected shape:

{
  "id": "chatcmpl-xxx",
  "choices": [
    {
      "message": {
        "role": "assistant",
        "content": "OK"
      }
    }
  ],
  "usage": {
    "prompt_tokens": 18,
    "completion_tokens": 1,
    "total_tokens": 19
  }
}

Discover models first

Before you ship fixed model behavior, query the model catalog:

curl "https://api.xenodia.xyz/v1/models"

/v1/models is public. Use this response to choose model IDs, supported modalities, async support, pricing mode, reference input rules, and available channels before making authenticated runtime calls.

Agent runtime quickstart

Use the CLI when you are setting up an agent environment rather than hand-writing HTTP calls:

npx xenodia@latest init

The setup path should detect wallet readiness, install Xenodia skill instructions, help locate or configure an API key, and point the agent back to these docs for full endpoint details.

Choose the operating mode explicitly:

For wallet-aware agents, bind the owner first, then create or assign the agent identity. The owner boundary decides who pays and which models or channels the agent may use.

Next steps

• Read Authentication for key handling.
• Read Chat Completions for the text endpoint.
• Read Image Generation for media workflows.
• Read Async Media Workflow before building polling.