# Chat Completions quickstart

Use a model ID returned by `GET /models` and send a Chat Completions request.

```bash
curl -X POST "$GENESIS_BASE_URL/chat/completions" \
  -H "Authorization: Bearer $GENESIS_ACCESS_TOKEN" \
  -H "Content-Type: application/json" \
  -d '{
    "model": "<CHAT_MODEL_ID>",
    "messages": [{"role": "user", "content": "Explain data lineage in one sentence."}],
    "max_tokens": 1024
  }'
```

Read the generated text from `choices[0].message.content`. Save the returned request ID and model ID when diagnosing a call.