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

# Regions

> Run Inworld in fully isolated regional deployments in the EU and India for data residency and lower local latency

In addition to our US deployment, Inworld operates fully isolated regional deployments in the **EU** and **India**. Regional deployments serve two needs:

* **Data residency** — your data is stored and processed in-region, supporting contractual and regulatory requirements such as GDPR data-localization commitments.
* **Latency** — inference runs close to your users, reducing round-trip time for latency-sensitive, real-time workloads.

<Note>
  Regional deployments are available through Enterprise contracts only — there is no self-serve signup. [Contact sales](https://inworld.ai/contact-sales) to get access.
</Note>

## How regional deployments work

Each region runs its own complete, independent Inworld stack — model inference, APIs, Portal, databases, and storage. Customer inputs (text, audio) and generated outputs are processed and stored entirely in-region.

Regions are isolated from one another:

* **Separate accounts and workspaces.** Your regional workspace starts empty — workspaces, API keys, and usage data do not carry over from the US region.
* **Voices are per-region.** A voice cloned or designed in one region is not available in another.
* **Regional deployments run in parallel with US.** The same organization can operate in multiple regions simultaneously — for example, serving EU end-users from the EU region while using the US region for everything else.

## Compliance

Regional data residency complements Inworld's existing security and privacy posture:

* **GDPR** — Inworld is GDPR-compliant; EU regional deployment additionally keeps processing and storage in the EU.
* **SOC 2 Type II** — applies across deployments.
* **[Zero Data Retention](/portal/zero-data-retention)** — eligible inputs and outputs are redacted from logs after processing; available in regional deployments and configurable per workspace.
* **HIPAA** — available with a BAA on Enterprise plans.

## Endpoints

Each region has its own API and Portal endpoints. WebSocket streaming connections use the same regional API host. Your existing US integration is unaffected — the default endpoints continue to work unchanged.

| Region       | API                 | Portal                   |
| :----------- | :------------------ | :----------------------- |
| US (default) | `api.inworld.ai`    | `platform.inworld.ai`    |
| EU           | `api.eu.inworld.ai` | `platform.eu.inworld.ai` |
| India        | `api.in.inworld.ai` | `platform.in.inworld.ai` |

<Note>
  API keys are region-specific. Create keys in the regional Portal ([platform.eu.inworld.ai](https://platform.eu.inworld.ai) or [platform.in.inworld.ai](https://platform.in.inworld.ai)) — keys from the US region will not authenticate against regional endpoints.
</Note>

## Product availability

| Product                                             | US                              | EU                                            | India                                         |
| :-------------------------------------------------- | :------------------------------ | :-------------------------------------------- | :-------------------------------------------- |
| [Text-to-Speech](/tts/tts)                          | <Icon icon="check" size={18} /> | <Icon icon="check" size={18} />               | <Icon icon="check" size={18} />               |
| [Instant Voice Cloning](/tts/instant-voice-cloning) | <Icon icon="check" size={18} /> | <Icon icon="check" size={18} />               | <Icon icon="check" size={18} />               |
| [Voice Design](/tts/voice-design)                   | <Icon icon="check" size={18} /> | <Icon icon="check" size={18} /> <sup>\*</sup> | <Icon icon="check" size={18} /> <sup>\*</sup> |
| [Speech-to-Text](/stt/overview)                     | <Icon icon="check" size={18} /> | <Icon icon="dash" size={18} />                | <Icon icon="dash" size={18} />                |
| [LLM Router](/router/introduction)                  | <Icon icon="check" size={18} /> | <Icon icon="dash" size={18} />                | <Icon icon="dash" size={18} />                |
| [Realtime API](/realtime/overview)                  | <Icon icon="check" size={18} /> | <Icon icon="dash" size={18} />                | <Icon icon="dash" size={18} />                |

<sup>\*</sup> AI-assisted voice description and preview script generation are not available in regional deployments — write your own description and script.

Want a product in a region where it isn't available yet? [Contact sales](https://inworld.ai/contact-sales) — regional product availability is driven by customer demand, and we are actively expanding regional coverage.

## Using a regional endpoint

Point your integration at the regional base URL — request and response formats, authentication, and API paths are identical across regions:

```bash theme={"system"}
curl -X POST "https://api.eu.inworld.ai/tts/v1/voice" \
  -H "Authorization: Basic $INWORLD_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "text": "Hello from the EU region!",
    "voiceId": "Ashley",
    "modelId": "inworld-tts-2"
  }'
```

The same applies to WebSocket streaming — connect to `wss://api.eu.inworld.ai` or `wss://api.in.inworld.ai` with the same paths you use today. If you use an SDK or tool that accepts a custom base URL, point it at the regional host.

## Limitations

* **No self-serve billing.** Usage is governed by your Enterprise contract; the Billing page is not available in regional Portals.
* **Access is provisioned for you.** Your organization is allowlisted during onboarding — sign up with your work email and your workspace is upgraded automatically.
* **LLM-powered conveniences are disabled.** Features that depend on hosted LLMs (for example, AI-generated voice descriptions in Voice Design) are hidden in regional Portals.
* **Model availability may vary by region.** Contact your account team for the current model list in your region.

## FAQ

<AccordionGroup>
  <Accordion title="Can I use the US region and a regional deployment in parallel?">
    Yes. Regional deployments are independent environments — many customers serve region-local traffic from the regional endpoint while continuing to use the US region for other workloads. Your application controls the routing.
  </Accordion>

  <Accordion title="Does using a regional deployment make me GDPR-compliant?">
    Regional deployment supports your compliance program by keeping storage and processing in-region, but compliance depends on your overall data-handling practices. Review your obligations with your legal team; [contact sales](https://inworld.ai/contact-sales) for our DPA and compliance documentation.
  </Accordion>

  <Accordion title="What latency improvement should I expect?">
    It depends on where your users are relative to your current endpoint. The regional deployment removes the intercontinental round trip for region-local users; model inference time is unchanged.
  </Accordion>

  <Accordion title="My API calls to a regional endpoint fail with an authentication error.">
    The most common cause is using a US-region API key. Keys are region-specific — create a key in the regional Portal and use it against the matching regional endpoint.
  </Accordion>
</AccordionGroup>

## Next Steps

<CardGroup cols={2}>
  <Card title="Contact Sales" icon="envelope" href="https://inworld.ai/contact-sales">
    Get access to the EU or India region for your organization.
  </Card>

  <Card title="Zero Data Retention" icon="shield-check" href="/portal/zero-data-retention">
    Learn how Inworld handles data privacy across TTS, STT, and LLMs.
  </Card>
</CardGroup>