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

# Overview

> The SaaS Gateway is a fully managed Guardway gateway — one click to set up, nothing to deploy, the same guardrails and controls.

The **SaaS Gateway** is a Guardway gateway that **we run for you**. One click in the dashboard provisions a private, dedicated gateway in Guardway's cloud; your applications point at a single public endpoint and everything else — infrastructure, upgrades, scaling the platform — is Guardway's job.

<Note>
  The SaaS Gateway is in **Beta**. It is fully functional and safe to use for real workloads, but you may see rough edges, and we may evolve details as we learn. Feedback is very welcome.
</Note>

## Self-hosted or SaaS?

Both are the same gateway, with the same guardrails, budgets, routing, and monitoring — the difference is who runs it and where your traffic flows.

<CardGroup cols={2}>
  <Card icon="box" title="Self-hosted gateway">
    Runs **in your network**. Prompts and completions never leave your infrastructure. You deploy and upgrade the container. Best when data locality is a hard requirement.
  </Card>

  <Card icon="cloud" title="SaaS Gateway (Beta)">
    Runs **in Guardway's cloud**, dedicated to your organization. Nothing to deploy or upgrade — ready in about two minutes. Best when you want the fastest path to a governed LLM endpoint.
  </Card>
</CardGroup>

<Note>
  **Data path difference, stated plainly:** with the SaaS Gateway your requests flow through Guardway-managed infrastructure on their way to your LLM provider. Your gateway is **dedicated and isolated** — its own private instance and its own database, never shared with other customers — but unlike self-hosted, traffic does transit our cloud. Request logs are retained for 30 days in the [Logs](/platform/logs) view.
</Note>

## How it works

Every organization gets its own gateway instance behind one shared front door:

1. Your app calls `https://gateway.guardway.ai` with your Guardway API key.
2. The edge identifies your organization from the key and routes the request to **your** dedicated gateway — never anyone else's.
3. Your gateway applies your guardrails, budgets, and routing, forwards the request to your LLM provider, and streams the response back.

Everything you configure in the dashboard — providers, API keys, guardrails, spend limits — applies to your SaaS Gateway automatically, usually within a minute.

## Good to know

|                           |                                                                   |
| ------------------------- | ----------------------------------------------------------------- |
| Region                    | United States                                                     |
| Endpoint                  | `https://gateway.guardway.ai` (OpenAI-compatible, `/v1`)          |
| Gateways per organization | One SaaS Gateway per organization                                 |
| Max request size          | 6 MB (larger requests get a fast `413`)                           |
| Streaming                 | Supported (SSE)                                                   |
| Request logs              | 30-day retention in the Logs view                                 |
| Access after deletion     | API keys stop routing within \~30 seconds of deleting the gateway |

## Where to go next

<CardGroup cols={2}>
  <Card icon="rocket" title="Set up your SaaS Gateway" href="/guardway-gateway/saas/setup">
    One click to provision, first request in minutes.
  </Card>

  <Card icon="arrows-rotate" title="Manage, redeploy, delete" href="/guardway-gateway/saas/manage">
    What each action does — and exactly what happens to your data.
  </Card>
</CardGroup>