> For the complete documentation index, see [llms.txt](https://studio.docs.ai.vonage.com/llms.txt). Markdown versions of documentation pages are available by appending `.md` to page URLs; this page is available as [Markdown](https://studio.docs.ai.vonage.com/ai-studio/create-a-new-agent.md).

# Create a new agent

The **Agents** tab allows you to create, modify, customize, and deploy Vonage AI virtual agents. Agents are presented in a box view by default. Each box displays key agent information at a glance.

You can create an unlimited number of agents under a single Vonage AI account and build multiple experiences in one place.

To create a virtual agent, click **Create agent** in the top-right corner of the screen.

{% hint style="info" %}
**Importing an existing agent**&#x20;

If you have previously exported an agent, click **Import agent** in the top-right corner. You can configure voice settings during the import flow. For more information, see [Voices](#voices).
{% endhint %}

## **Step 1:** Select the agent type

After clicking **Create agent**, select the agent type that matches your use case:

* [<mark style="color:purple;">Telephony</mark>](https://studio.docs.ai.vonage.com/voice/get-started) for voice-based flows.
* [<mark style="color:purple;">WhatsApp</mark>](https://studio.docs.ai.vonage.com/whatsapp-chatbot/whatsapp-chatbot) for WhatsApp-integrated chatbots.
* [<mark style="color:purple;">SMS</mark>](https://studio.docs.ai.vonage.com/whatsapp-chatbot/sms-chatbot-coming-soon) for SMS-based chatbots.
* [<mark style="color:purple;">HTTP</mark>](https://studio.docs.ai.vonage.com/api-integration/working-with-http-agents) for web-based chatbots.

## Step 2: Select the AI engine

Select the AI engine that determines how your agent processes conversations. The following options are available:

| AI engine   | Availability           | Description                                                                                                                                                                                                                                                              |
| ----------- | ---------------------- | ------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------ |
| Agentic     | Gated feature          | Operates autonomously using the AI Agentic framework. Makes context-based decisions to achieve specific goals with minimal human intervention. For more information, see [Agentic NLU AI engine.](/ai-studio/agentic-nlu-ai-engine.md)                                   |
| Hybrid      | Available to all users | Combines the contextual understanding of large language models with structured natural language processing. Delivers improved overall performance. For more information, see [NLU AI engine - Traditional vs Hybrid](/ai-studio/nlu-ai-engine-traditional-vs-hybrid.md). |
| Traditional | Available to all users | Uses statistical models for rule-driven intent recognition and entity extraction. Provides high predictability but less flexibility. For more information, see [NLU AI engine - Traditional vs Hybrid](/ai-studio/nlu-ai-engine-traditional-vs-hybrid.md).               |

## Step 3: Add agent information

Fill in the following fields:

<table data-search="false"><thead><tr><th>Field</th><th>Options</th><th>Description</th></tr></thead><tbody><tr><td>Region</td><td><ul><li>United States.</li><li>Europe.</li></ul></td><td>Select whether your agent operates in the US or in Europe. The selected region determines the agent's response time.</td></tr><tr><td>Agent Name</td><td>Text</td><td>Enter a unique name that reflects the agent's purpose. You can change the name at any time, and it is visible only to you.</td></tr><tr><td>API Key</td><td>A list of available API keys.</td><td>Displays your Vonage account or sub-account key. Use this key to associate a phone number with your agent via the API dashboard.</td></tr><tr><td>Language</td><td>A list of available languages.</td><td>Select the language your agent will use. After you select a language, you cannot change it for that agent. For a full list of supported languages, see <a href="/theres-more/languages-available.md">Languages available</a>. </td></tr><tr><td>Voices (voice agents only)</td><td><ul><li>Standard voice.</li><li>Premium voice.</li><li>Premier voice.</li></ul></td><td>Select the voice type and voice for your agent. For more information, see <a href="#voices">Voices</a>.</td></tr><tr><td>Time Zone</td><td>A list of time zones.</td><td>Select the time zone in which your agent will operate. All reports are generated according to the selected time zone. You can change the time zone at any time.</td></tr></tbody></table>

## Step 4: Select a template or start from scratch

Select a ready-made template agent to use as a starting point, or start with a blank flow. If you select a template, you can modify and customise it to fit your needs.

## Step 5: Select the event type

Select whether your agent will handle **inbound calls** (the agent receives calls) or **outbound calls** (the agent initiates calls). You can add additional event types to the agent later.

For more information, see

* [Voice events](/voice/events.md).
* [WhatsApp events](/whatsapp/events.md).
* [SMS events](/sms/events.md).
* [HTTP events](/http/events.md).

## Step 6: Build your conversational flow

Your agent is now created. Begin building your conversational flow node by node.

For step-by-step guidance, see the get-started guides for [<mark style="color:purple;">Voice</mark>](https://studio.docs.ai.vonage.com/voice/get-started/create-your-first-conversational-flow), [<mark style="color:purple;">WhatsApp</mark>](https://studio.docs.ai.vonage.com/whatsapp/get-started/create-your-first-conversational-flow), [<mark style="color:purple;">SMS</mark>](https://studio.docs.ai.vonage.com/sms/get-started/create-your-first-conversational-flow)<mark style="color:purple;">,</mark> and [<mark style="color:purple;">HTTP</mark>](https://studio.docs.ai.vonage.com/http/working-with-http-agents/create-your-first-conversational-flow) agents.&#x20;

![](https://3877181490-files.gitbook.io/~/files/v0/b/gitbook-x-prod.appspot.com/o/spaces%2F-L_81A0PNZfdawu_TPAO%2Fuploads%2Fi5OtZpbZdAgugnJxFjni%2FScreen%20Shot%202022-05-29%20at%2018.36.43.png?alt=media\&token=b7eb6f87-96bd-4a1d-a237-d3c9e68fd7b2)

## Voices

Vonage AI Studio offers three voice types for voice agents: **Standard voice**, **Premium voice**, and **Premier voice**. The voice type you select determines the text-to-speech engine used and the audio quality your callers hear.

After you select a voice, Vonage AI Studio applies it to all Speak nodes, Collect Input nodes, Agentic Capture nodes, and Agentic Classification nodes within the agent.&#x20;

You can also update the voice at any time from **Agent Details**, or during the **Import agent** and **Duplicate agent** flows.

### Standard and Premium voices

Standard and Premium voices are available to all users by default. Both are powered by the [Vonage Text-to-Speech service](https://developer.vonage.com/en/voice/voice-api/concepts/text-to-speech?source=voice#standard-and-premium). Premium voice produces higher audio quality than the Standard voice.

Both voice types support Speech Synthesis Markup Language (SSML) tags, which allow you to control speech delivery programmatically.

### Premier voice

{% hint style="warning" %}
**Gated feature**

Premier voices are currently a gated feature. For more information, please contact Vonage Support or your Account Manager.
{% endhint %}

Premier voice uses generative AI to produce more natural, human-like speech. It is powered by [Google Cloud Text-to-Speech (Chirp3 HD)](https://docs.cloud.google.com/text-to-speech/docs/chirp3-hd). For scripting best practices, refer to the [Google Cloud Text-to-Speech documentation](https://docs.cloud.google.com/text-to-speech/docs/chirp3-hd).

#### Punctuation techniques for Premier voice

Premier voice does not support SSML tags. Use punctuation marks to control pacing and flow in your scripts instead. The following punctuation marks are supported:

| Punctuation mark | Effect                                                                                                     | Example                                                                  |
| ---------------- | ---------------------------------------------------------------------------------------------------------- | ------------------------------------------------------------------------ |
| Period (`.`)     | Full stop; creates a longer pause. Use to separate complete thoughts and define clear sentence boundaries. | "Your order has been placed. A confirmation will be sent to your email." |
| Comma (`,`)      | Short pause within a sentence. Use to separate clauses, list items, or introduce brief breaks.             | "Please hold, and I will transfer you to an agent."                      |
| Ellipsis (`...`) | Longer, deliberate pause. Use to indicate trailing thoughts, hesitation, or a dramatic pause.              | "We are locating your account... please stay on the line."               |
| Hyphen (`-`)     | Brief pause or a sudden break in thought.                                                                  | "To confirm - press 1. To cancel - press 2."                             |


---

# Agent Instructions
This documentation is published with GitBook. GitBook is the documentation platform designed so that both humans and AI agents can read, navigate, and reason over technical content effectively. Learn more at gitbook.com.

## Querying This Documentation
If you need additional information that is not directly available in this page, you can query the documentation by asking a question.

Perform an HTTP GET request on the following URL with the `ask` and `goal` query parameters:

```
GET https://studio.docs.ai.vonage.com/ai-studio/create-a-new-agent.md?ask=<question>&goal=<user_goal>
```

`ask` is the immediate question: it should be specific, self-contained, and written in natural language.
`goal` is what the user is ultimately trying to achieve, the reason they need the answer. Sharing it helps GitBook give you a better, more relevant answer. A goal is most helpful when it describes the outcome the user wants rather than restating the question. For example, with `ask=how do I create an API token`, a goal like `automate deployments from our CI pipeline` lets GitBook tailor the answer to that use case.

The response will contain a direct answer to the question and relevant excerpts and sources from the documentation.

Use this mechanism when the answer is not explicitly present in the current page, you need clarification or additional context, or you want to retrieve related documentation sections.
Field	Options	Description
Region	United States. Europe.	Select whether your agent operates in the US or in Europe. The selected region determines the agent's response time.
Agent Name	Text	Enter a unique name that reflects the agent's purpose. You can change the name at any time, and it is visible only to you.
API Key	A list of available API keys.	Displays your Vonage account or sub-account key. Use this key to associate a phone number with your agent via the API dashboard.
Language	A list of available languages.	Select the language your agent will use. After you select a language, you cannot change it for that agent. For a full list of supported languages, see Languages available.
Voices (voice agents only)	Standard voice. Premium voice. Premier voice.	Select the voice type and voice for your agent. For more information, see Voices.
Time Zone	A list of time zones.	Select the time zone in which your agent will operate. All reports are generated according to the selected time zone. You can change the time zone at any time.