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

# GravixLayer sandboxes

> Run CostGraph workflows and CI jobs in GravixLayer sandboxes: the key to create, its permissions, and its limits

Gravix Layer Agent Runtime gives each runtime its own sandbox with dedicated CPU, memory and disk, for running code, shell commands, a browser or a desktop. It is called over HTTPS.

The following table shows what CostGraph runs on this provider.

| Workflows | CI runners | Connect with |
| - | - | - |
| Yes | Yes | Integrations or your own dispatcher |

## Connect

Connect GravixLayer on Integrations, and CostGraph starts sandboxes in your GravixLayer account. You don't run anything yourself.

1. In CostGraph, go to **Integrations > Sandbox > GravixLayer**.

2. Enter these fields:

   * **API key** (`GRAVIXLAYER_API_KEY`, required): A Gravix Layer API key from the Gravix Layer console.

3. Click **Connect**.

Sandbox size comes from the template set as `providerTemplate`.

### Your own dispatcher

To keep the key out of CostGraph or reach a private network, run your own dispatcher instead.
Set `provider` to `gravixlayer` and the preceding fields as environment variables, as described in [Run your own dispatcher](https://docs.costgraph.ai/costgraph/workflows/managed-sandboxes.md#run-your-own-dispatcher).

## Key and permissions

GRAVIXLAYER\_API\_KEY is a Gravix Layer API key. Create it in the console at app.gravixlayer.ai; it is shown once. It is sent as a Bearer token.

The provider doesn't document key scopes or roles, so a key has full access to the account.

Sources: [docs.gravixlayer.ai: sdk usage](https://docs.gravixlayer.ai/documentation/agentruntime/sdk-usage.md), [docs.gravixlayer.ai: runtime overview](https://docs.gravixlayer.ai/documentation/agentruntime/runtime-overview.md), [docs.gravixlayer.ai: templates](https://docs.gravixlayer.ai/documentation/agentruntime/templates.md), [docs.gravixlayer.ai: overview](https://docs.gravixlayer.ai/documentation/agentruntime/overview.md), [docs.gravixlayer.ai: update timeout](https://docs.gravixlayer.ai/documentation/agentruntime/sandbox-management/update-timeout.md).

## Limits

* The runtime terminates when it reaches its timeout, which is set in seconds and can be extended. The SDK example uses 300 s. No maximum is documented. ([source](https://docs.gravixlayer.ai/documentation/agentruntime/sandbox-management/update-timeout.md))
* Pause frees CPU and memory and keeps the runtime. Automatic idle pause is not documented. ([source](https://docs.gravixlayer.ai/documentation/agentruntime/runtime-overview.md))
* Template sizes: base-small 1 vCPU/1 GB/4 GB disk, base-medium 2/2 GB/6 GB, base-large 4/4 GB/8 GB. No larger maximum is stated. ([source](https://docs.gravixlayer.ai/documentation/agentruntime/templates.md))
* Regions: AWS us-east-1 (default) and AWS ap-south-1. The docs do not state the CPU architecture. ([source](https://docs.gravixlayer.ai/documentation/agentruntime/overview.md))

## Related

* [Supported sandbox providers](https://docs.costgraph.ai/costgraph/workflows/sandbox-providers.md)
* [Run workflows on a managed sandbox provider](https://docs.costgraph.ai/costgraph/workflows/managed-sandboxes.md)
* [Run CI jobs in sandboxes](https://docs.costgraph.ai/costgraph/autopilot/ci-sandboxes.md)


This documentation is built and hosted on [Mintlify](https://mintlify.com), a developer documentation platform.