> For the complete documentation index, see [llms.txt](https://anunzi-ai.gitbook.io/anunzi-docs/llms.txt). Markdown versions of documentation pages are available by appending `.md` to page URLs; this page is available as [Markdown](https://anunzi-ai.gitbook.io/anunzi-docs/api-widgets/trabajar-con-la-api-de-llamadas/assistants/create-assistant.md).

# Crear agente

**POST**\
`https://api.anunzi.net/v2/assistants`

Create a new AI call assistant in your Anunzi environment. Use this endpoint to configure inbound or outbound assistants, setting their behavior, voice, and connection details.

***

#### Body Parameters

| Parameter                | Type    | Required | Description                                                                |
| ------------------------ | ------- | -------- | -------------------------------------------------------------------------- |
| `type`                   | string  | Yes      | The type of assistant to be created. Options: `"inbound"` or `"outbound"`. |
| `name`                   | string  | Yes      | Name of the person who will be called.                                     |
| `phone_number`           | string  | No       | Retrieve your phone numbers using the "Get Numbers" API endpoint.          |
| `caller_id_number`       | string  | No       | Phone number to be displayed as caller ID (e.g., +12345678901).            |
| `external_webhook_url`   | string  | No       | URL to receive post-call data (transcripts, etc.).                         |
| `is_recording`           | boolean | No       | Enable to record conversations, accessible in the calls section.           |
| `is_transcript_disabled` | boolean | No       | Defaults to `false`. Disable to prevent transcription of conversations.    |
| `agent`                  | object  | Yes      | Agent configuration object (see details below).                            |

***

#### Agent Object

| Parameter           | Type    | Required | Description                                                       |
| ------------------- | ------- | -------- | ----------------------------------------------------------------- |
| `prompt`            | string  | Yes      | Dynamic prompt using variables (e.g., `{your-variable}`).         |
| `greeting_message`  | string  | Yes      | Greeting message with variable support (e.g., `{your-variable}`). |
| `llm`               | string  | Yes      | Defaults to `"anunzi"`. Determines the LLM used by the assistant. |
| `language`          | string  | Yes      | Language the assistant will use to speak (e.g., `"en-US"`).       |
| `voice_id`          | string  | Yes      | Retrieve available voices using the "Get Voices" API endpoint.    |
| `patience_level`    | string  | No       | Defaults to `"low"`. Options: `"low"`, `"medium"`, `"high"`.      |
| `timezone`          | string  | No       | Defaults to `"Europe/Amsterdam"`.                                 |
| `consent_recording` | boolean | No       | If enabled, the agent will ask for consent to record the call.    |
| `consent_text`      | string  | No       | The message the agent will use to request recording consent.      |

***

#### Example Request (cURL)

```bash
curl --request POST \
     --url https://api.anunzi.net/v2/assistants \
     --header 'accept: application/json' \
     --header 'content-type: application/json' \
     --header 'Authorization: Bearer API_KEY' \
     --data '
{
  "type": "inbound",
  "agent": {
    "llm": "anunzi",
    "language": "en-US"
  },
  "name": "test",
  "phone_number": "+54898563231"
}
'
```

***

#### Example Response (200 OK)

```json
{
  "status": "ok",
  "response": {
    "model_id": "1726559381711x108708587477942990"
  },
  "details": {
    "phone": "",
    "voice": ""
  }
}
```

***

#### Status Codes

| Code | Description                                 |
| ---- | ------------------------------------------- |
| 200  | Assistant created successfully.             |
| 400  | Bad Request. Invalid or missing parameters. |