# Quickstart

> Create a Cascadia account, add your first device and send it a request.

To get started, you'll need a machine that meets the following [requirements](/enterprise/getting-started/requirements/) and has administrator privileges.

## 1. Create your account

<Steps>

1. Go to [console.cascadia.to/signup](https://console.cascadia.to/signup) and create an account with your work email.

2. Open the confirmation link in your email, and sign in.

3. Name your organization and choose its **URL name**. Your fleet's certificate authority lives at `ca.cascadia.to/o/<url-name>`.

</Steps>

:::caution
The URL name is written into every device certificate, so it can't be changed later.
:::

If a colleague already set up your organization, ask them to invite you from **Members** instead of creating a new one.

## 2. Add your first device

<Steps>

1. In the console, open **Devices → Add device**. When it asks who will run the installer, choose **I will**.

2. Select **Create installer**, then **Download installer** for your operating system.

3. Run the installer on the PC and accept the prompts. On Windows these are **Keep**, **Run** and **Yes**; on Ubuntu, **Install** and your password.

4. Wait for the install to finish (this typically takes 2 to 10 minutes). It sets up the node software and a small model, then enrolls the PC in your fleet.

</Steps>

The device appears on the console's **Devices** page as soon as it enrolls.

If the install stops, see [Installing a device](/enterprise/troubleshooting/#installing-a-device).

## 3. Create an API token

<Steps>

1. In the console, open **Settings → API tokens**.

2. Enter a label, such as `quickstart`, choose an expiry and select **Create token**.

3. Copy the token. The console shows it only once.

</Steps>

## 4. Send a request

On the device you just added, list the models it can reach, then send a chat request with one of their IDs. Replace `<TOKEN>` with your token and `<MODEL_ID>` with an `id` from the list.

**Windows**

In PowerShell:

```powershell
$headers = @{ Authorization = "Bearer <TOKEN>" }
Invoke-RestMethod http://127.0.0.1:9100/v1/models -Headers $headers

$body = '{"model":"<MODEL_ID>","messages":[{"role":"user","content":"Hello!"}]}'
Invoke-RestMethod http://127.0.0.1:9100/v1/chat/completions -Method Post `
  -Headers $headers -ContentType "application/json" -Body $body
```

**Ubuntu**

In a terminal:

```bash
curl http://127.0.0.1:9100/v1/models \
  -H "Authorization: Bearer <TOKEN>"

curl http://127.0.0.1:9100/v1/chat/completions \
  -H "Authorization: Bearer <TOKEN>" \
  -H "Content-Type: application/json" \
  -d '{"model":"<MODEL_ID>","messages":[{"role":"user","content":"Hello!"}]}'
```

The gateway only accepts requests from the device it runs on. Your fleet then routes each request to whichever devices serve the model you asked for.

## Next steps

- [Add more devices](/enterprise/devices/add-a-device/) to your network (one at a time or in a batch)
- [Install the Fleet Manager](/enterprise/fleet-manager/install/) to deploy larger models
- [Connect an application](/enterprise/use/connect-an-application/)
- [Invite team members](/enterprise/use/team-and-roles/) to your Cascadia organization
