Quickstart

Five steps from zero to a running agent. Every command below works unmodified against the production API once you substitute your own key and agent ID.

1. Generate an API key

Go to Settings → API Keys → Create Key (see Authentication) and copy the key. It's shown once.

2. Verify your key

curl -H "X-API-Key: elz_your_key_here" \
  https://api.elizon.com/api/v1/organizations/limits

{
  "rate_limits": { "org_rpm": 1000, "key_rpm": null },
  "file_uploads": {
    "enabled": false,
    "max_file_size_mb": 10,
    "max_total_size_mb": 50,
    "max_files_per_request": 20,
    "note": "File upload support is not yet enabled in this version."
  }
}

A 200 with this shape confirms your key is valid.

3. List your agents

curl -H "X-API-Key: elz_your_key_here" \
  https://api.elizon.com/api/v1/agents

{
  "data": [
    {
      "agentId": "3f9c...",
      "name": "Support Bot",
      "status": "active",
      "projectId": null,
      "...": "..."
    }
  ],
  "total": 1,
  "page": 1,
  "limit": 20
}

Copy an agentId from the response — you'll use it in the next step.

4. Run an agent

curl -X POST https://api.elizon.com/api/v1/agents/{agentId}/run \
  -H "X-API-Key: elz_your_key_here" \
  -H "Content-Type: application/json" \
  -d '{"message": "Hello!"}'

5. Read the response

{
  "runId": "b7e1...",
  "agentId": "3f9c...",
  "agentName": "Support Bot",
  "content": "Hello! How can I help you today?",
  "sessionId": "a1c4...",
  "userId": "d92e...",
  "createdAt": 1732550123,
  "success": true,
  "metrics": {
    "inputTokens": 12,
    "outputTokens": 9,
    "latencyMs": 842,
    "model": null,
    "failoverUsed": false
  },
  "messages": null
}

content — the agent's reply.
sessionId — pass this back on your next call to the same agent to continue the conversation; omit it to start a new one.
metrics — token counts and latency for this run. model is currently always null.

That's the whole loop. From here:

Need the response as it's generated, token by token? See POST /run-stream.
Running something that might take longer than 30 seconds? See POST /run-async.
Want to create or update agents instead of just running them? See the API Reference.
Composing several agents together? See Teams. Chaining steps into a graph? See Workflows.