Skip to navigation

Quickstart

Build your first conversational agent in as little as 5 minutes.

In this guide, you’ll learn how to create your first conversational agent. This will serve as a foundation for building conversational workflows tailored to your business use cases. Let your coding agent set it up, or follow the steps below by hand.

Set up with your coding agent

You do not need an account or an API key before you start. Paste this into your coding agent:

Add ElevenLabs to my project: run `npx -y @elevenlabs/cli@latest onboard init` and follow what it prints.

The ElevenLabs CLI takes it from there and tells your coding agent each step. Your browser opens once so you can sign in or create an account and click Authorize. The key is saved in your project’s env file (.env.local, or .env in a Python project) and kept out of git; it is never shown in the chat. When your coding agent asks what you’d like to build, describe the agent you want: what it should do, and where people will talk to it, such as a widget on your website or a voice button in your app. Your coding agent then creates it, adds it to your app, and gives you a link to try it.

If your coding agent runs somewhere your browser cannot reach, such as a remote machine over SSH or a cloud sandbox, the CLI asks you to create a key yourself and paste it into the env file, then continues. The command needs Node.js for npx. Without it, install the ElevenLabs CLI with Homebrew (macOS and Linux) or Scoop (Windows), or update it if you already have it (brew upgrade elevenlabs, scoop update elevenlabs), and ask your coding agent to run elevenlabs onboard init instead.

To keep the ElevenLabs agents skill in your coding agent for later work on your agents, install it:

npx skills add elevenlabs/skills --skill agents

To build agents from Claude, ChatGPT or another MCP client instead, connect the hosted MCP server.

Getting started

ElevenLabs Agents are managed either through the ElevenAgents dashboard, the ElevenLabs API, the Agents CLI or the hosted MCP server.

ElevenLabs Agents

The assistant at the bottom right corner of this page is an example of an ElevenLabs agent, capable of answering questions about ElevenLabs, navigating pages & taking you to external resources.

Creating your first agent

In this quickstart guide we’ll start by creating an agent via the API or the web dashboard. Next we’ll test the agent, either by embedding it in your website or via the ElevenLabs dashboard.

In this guide, we’ll create a conversational support assistant capable of answering questions about your product, documentation, or service. This assistant can be embedded into your website or app to provide real-time support to your customers.

ElevenLabs Agents

The assistant at the bottom right corner of this page is capable of answering questions about ElevenLabs, navigating pages & taking you to external resources.
1

Sign in to ElevenLabs

Go to elevenlabs.io and sign in to or create your account.

2

Create a new assistant

In the ElevenLabs Dashboard, create a new assistant by entering a name and selecting the Blank template option.

Dashboard

Creating a new assistant
3

Configure the assistant behavior

Go to the Agent tab to configure the assistant’s behavior. Set the following:

1

First message

This is the first message the assistant will speak out loud when a user starts a conversation.

First message
Hi, this is Alexis from <company name> support. How can I help you today?
2

System prompt

This prompt guides the assistant’s behavior, tasks, and personality.

Customize the following example with your company details:

System prompt
You are a friendly and efficient virtual assistant for [Your Company Name]. Your role is to assist customers by answering questions about the company's products, services, and documentation. You should use the provided knowledge base to offer accurate and helpful responses.
Tasks:
- Answer Questions: Provide clear and concise answers based on the available information.
- Clarify Unclear Requests: Politely ask for more details if the customer's question is not clear.
Guidelines:
- Maintain a friendly and professional tone throughout the conversation.
- Be patient and attentive to the customer's needs.
- If unsure about any information, politely ask the customer to repeat or clarify.
- Avoid discussing topics unrelated to the company's products or services.
- Aim to provide concise answers. Limit responses to a couple of sentences and let the user guide you on where to provide more detail.
4

Add a knowledge base

Go to the Knowledge Base section to provide your assistant with context about your business.

This is where you can upload relevant documents & links to external resources:

  • Include documentation, FAQs, and other resources to help the assistant respond to customer inquiries.
  • Keep the knowledge base up-to-date to ensure the assistant provides accurate and current information.

Next we’ll configure the voice for your assistant.

1

Select a voice

In the Voice tab, choose a voice that best matches your assistant from the voice library:

Voice settings

Using higher quality voices, models, and LLMs may increase response time. For an optimal customer experience, balance quality and latency based on your assistant’s expected use case.
2

Testing your assistant

Press the Test AI agent button and try conversing with your assistant.

Configure evaluation criteria and data collection to analyze conversations and improve your assistant’s performance.

1

Configure evaluation criteria

Navigate to the Analysis tab in your assistant’s settings to define custom criteria for evaluating conversations.

Analysis settings

Every conversation transcript is passed to the LLM to verify if specific goals were met. Results will either be success, failure, or unknown, along with a rationale explaining the chosen result.

Let’s add an evaluation criteria with the name solved_user_inquiry:

Prompt
The assistant was able to answer all of the queries or redirect them to a relevant support channel.
Success Criteria:
- All user queries were answered satisfactorily.
- The user was redirected to a relevant support channel if needed.
2

Configure data collection

In the Data collection section, configure details to be extracted from each conversation.

Click Add item and configure the following:

  1. Data type: Select “string”
  2. Identifier: Enter a unique identifier for this data point: user_question
  3. Description: Provide detailed instructions for the LLM about how to extract the specific data from the transcript:
Prompt
Extract the user's questions & inquiries from the conversation.
Test your assistant by posing as a customer. Ask questions, evaluate its responses, and tweak the prompts until you’re happy with how it performs.
3

View conversation history

View evaluation results and collected data for each conversation in the Call history tab.

Conversation history

Regularly review conversation history to identify common issues and patterns.

The newly created agent can be tested in a variety of ways, but the quickest way is to use the ElevenLabs dashboard.

The web dashboard uses our React SDK under the hood to handle real-time conversations.

If instead you want to quickly test the agent in your own website, you can use the Agent widget. Simply paste the following HTML snippet into your website, taking care to replace agent-id with the ID of your agent.

<elevenlabs-convai agent-id="agent-id"></elevenlabs-convai>
<script src="https://unpkg.com/@elevenlabs/convai-widget-embed" async type="text/javascript"></script>

Next steps

As a follow up to this quickstart guide, you can make your agent more effective by integrating: