Skip to main content
You can also build Voice Agents on Line from the playground by simply describing what you want the Agent. You can export the code and customize it even more.
1

Sign into Cartesia Playground

Go to the Text to Agent page in the playground.
2

Describe your Agent

Enter a description of what you want the Agent to accomplish.
Select a template
3

Try out candidates

After a generation run completes, you can try calling several candidates created and pin one to production. This will assign it to a phone number and count towards your Agent slots. You can also re-generate a new version of your Agent by repeating the steps above.
Try canddiates
4

Export code to GitHub

If you want to further customize your Agent, you can export your Agent code to a new GitHub repository, modify it, and then re-deploy it. Start by connecting your Agent to GitHub in the playground.
Connect to Github
5

Clone your Agent locally

Now that your repo is created, let’s set up your agent on your machine.
git clone https://github.com/your-username/your-repo-name.git
6

Customize your Agent

Once that your repo is cloned, you can customize it with specific instructions in the README.md. You can use cartesia chat from your CLI to try your changes locally before deploying.
7

Deploy your changes

When you’re satisfied with your Agent, pushing to main will trigger a new deployment. zsh git add . git commit -m "hello world" git push
8

Check deployment status

See your deployment status in the playground or the CLI if you prefer.
Deployments

Text to Agent Demo