---
title: "Get started"
description: "Make an account and an API key, install the glac command line or an SDK, and take a first step with each product."
---

> Documentation Index
> Fetch the complete documentation index at: https://glac.ai/docs/llms.txt
> Use this file to discover all available pages before exploring further.

# Get started

## 1. An account and a key

Sign up in the [console](https://console.glac.ai/console/), and confirm your address from the email it sends: until then you can look, but not make anything. A new account is on the Free plan, with a first project called `default`. Open the project's **API keys** page and make a key. The key (`ak_…`) is shown once: keep it somewhere safe.

A key belongs to one project and has scopes. A key made without chosen scopes has every scope but `admin`. That is enough for the Cube, Sandbox and Drive quickstarts; storing secrets needs a key with `admin` too. See [projects and keys](/docs/platform/projects/).

Every tool reads the same two environment variables:

```bash
export GLAC_API_URL=https://api.glac.ai
export GLAC_API_KEY=ak_...
```

## 2. The tools

The command line, `glac`, is one file:

```bash
curl -fsSL https://glac.ai/install.sh | sh
glac version
```

Instead of the variables, `glac login --api-key -` reads the key from standard input and keeps it in `~/.config/glac/config.json`.

The Python SDK needs Python 3.9 or later and nothing else:

```bash
pip install glac
```

There are SDKs for [TypeScript](/docs/sdks/typescript/), [Go](/docs/sdks/go/), [Rust](/docs/sdks/rust/) and [Java](/docs/sdks/java/) too, and an [MCP server](/docs/mcp/) that gives an AI agent the products as tools.

## 3. A first step with each product

**[Sandbox](/docs/sandbox/quickstart/)**: run one command in a machine of its own.

```bash
glac run --image python:3.12 -- python3 -c 'print(6*7)'
```

**[Cube](/docs/cube/quickstart/)**: serve an image at an address of its own.

```bash
glac deploy public.ecr.aws/docker/library/python:3.12-alpine --name hello --port 8000 \
  --start-command '["python3", "-m", "http.server", "8000"]'
```

**[Drive](/docs/drive/quickstart/)**: a disk that outlives its sandbox (one Drive may be up to 10 GiB on Free).

```bash
glac drives create --name data --size-bytes 1GiB
glac sandboxes create --name box --drive data:/data
```

Every command takes `--json` and then prints what the API returned, and `glac help` lists the commands by product. When you are done, `glac sandboxes rm box`, `glac drives rm data` and `glac cube rm hello` delete what these made.

## Next

- [Pricing and plans](/docs/platform/pricing/): what Free includes and what usage past it costs.
- [Limits](/docs/platform/limits/): what each plan lets an account have at once.
- [API reference](/docs/api/): every call.

Source: https://glac.ai/docs/quickstart/index.mdx