- Getting started
- Data security and compliance
- Organizations
- Authentication and security
- Licensing
- About licensing
- Unified Pricing: Licensing plan framework
- Flex: Licensing plan framework
- Activating your Enterprise license
- Upgrading and downgrading licenses
- Requesting a service trial
- Assigning licenses to tenants
- Assigning user licenses
- Deallocating user licenses
- Monitoring license allocation
- License overallocation
- Licensing notifications
- User license management
- Tenants and services
- Accounts and roles
- Testing in your organization
- AI Trust Layer
- External applications
- Notifications
- Logging
- Troubleshooting
- Migrating to Automation Cloud

Automation Cloud admin guide
Some UiPath services use AI-powered capabilities delivered through third-party large language models (LLMs), including the GPT family (via Azure OpenAI Service), Gemini models (via Google Vertex AI), and Claude models (via Amazon Bedrock).
These models are hosted in specific regions. When an LLM endpoint is not available in the same region as your UiPath organization or tenant, data is temporarily routed to a region with an available model endpoint.
UiPath does not control the geographic availability of third-party LLMs. However, once a third-party provider makes a new model region available, UiPath investigates and works toward supporting that region for its AI-powered features, while continuing to define only the routing behavior.
When you use an AI feature, the data path depends on three separate regions:
| Region types | Description | Who controls it |
|---|---|---|
| Organization region |
The region where your UiPath organization is provisioned. The organization region is considered for services enabled at the organizational level, that inherit the organization region (for example, Apps, Automation Ops, and Studio Web). |
You choose the organization region when you create the organization. An organization region is not visible in the UI. |
| Tenant region | The region where a tenant-level service (for example, Orchestrator, Test Manager, or Agents) is hosted and processes data. |
You choose the tenant region when you create the tenant. A tenant’s region is visible in the UI, from the tenant dropdown. |
| LLM processing region | The region where the third-party model actually runs. | Determined by the third-party LLM provider, which defines where an LLM endpoint is available. If an endpoint is not available in the organization or tenant region, UiPath may temporarily route the request to another supported region. When a new LLM region becomes available, UiPath evaluates and works toward supporting it, while continuing to define only the routing behavior. |
These regions can differ. While your UiPath data remains in your selected organization or tenant region, data used when working with AI features may be temporarily routed to another region for model processing if no local LLM endpoint exists.
Check the following definitions to understand the residency scopes used when working with UiPath’s AI features:
-
Organization region: The geographic region you select when creating your UiPath organization. This determines where your organization data is stored, including the data for services that inherit the organization region. Such services are marked as Organization-level in the Global cloud regions table.
-
Tenant region: The region where a tenant is hosted and where its services (for example, Orchestrator, Test Manager, or Agents) operate. Services that inherit the tenant region are marked as Tenant-level in the Global cloud regions table.
-
Service region: The region where a specific service within an organization or tenant runs. In most cases, all services in a tenant follow the tenant region, but some follow the organization region. The data residency of a service is marked in the Global cloud regions table as either inheriting the organization region or tenant region, along with its availability across cloud regions.
-
LLM model endpoint region: The region where the third-party large language model (LLM) executes data processing. The availability of LLM regions is determined by the third-party provider (for example, Azure OpenAI, Google Gemini, or Anthropic Claude). UiPath defines routing rules only when a model is unavailable in the selected tenant or organization region.
The following example shows how data is handled across regions when using UiPath services that include AI features. It highlights the different residency scopes involved, and explains how data is processed.
| Scenario | Description |
|---|---|
| Organization region | European Union (EU) |
| Tenant region | EU |
| Services used |
|
| Model availability |
|
| LLM processing region | The LLM processing region is the same as the organization and tenant region: EU. All data is hosted and processed in the same region. |
The availability of a model endpoint in a given region is determined by the third-party model provider. UiPath only defines the region to which data is routed temporarily.
- AU - Australia
- CA - Canada
- CH - Switzerland
- EU - European Union
- IN - India
- JA - Japan
- SI - Singapore
- UAE - United Arab Emirates
- UK - United Kingdom
- US - United States
| Service | Model type and version | Endpoint region availability | Routing behavior | Notes |
|---|---|---|---|---|
| Apps | Azure OpenAI | CA, EU, US | For the AU, IN, and SI regions, data is re-routed to EU. | N/A |
| Agents | GPT 4o-x | EU, JA, US |
| N/A |
| Agents | GPT 4.1x | AU, CA, EU, IN, JA, SI, UK, US | Routed to the closest supported region. | N/A |
| Agents | GPT 5 and GPT 5 mini (Preview) | EU, US |
| N/A |
| Agents | Claude 3.5/3.7 |
|
| N/A |
| Agents | Gemini 2.5 Flash, Gemini 2.5 Pro | EU, US | Routed to the closest supported region. | N/A |
| Autopilot | Azure OpenAI (gpt-4o-mini-2024-07-18 and gpt4o-2024-08-06) | EU, JA, US (Autopilot for Developers),AU, CA, EU, IN, JA, UK, US (Autopilot for Testers) |
| N/A |
| Autopilot | Google Gemini (gemini-2.0-flash-001 and gemini-2.5-flash) | EU, US |
| N/A |
| Autopilot | Anthropic Claude (claude-3-7-sonnet-20250219-v1:0) | EU, US |
| N/A |
| Autopilot for Everyone (fast model) | GPT-4o-mini-2024-07-18 (Azure OpenAI) | AU, CA, EU, IN, JA, SI, UK, US |
|
Autopilot for Everyone uses GPT-4o-mini as its fast model for smaller tasks. Autopilot for Everyone stores customer data in Orchestrator tenant region. Autopilot for Everyone backend service (data transit only) is deployed for the following regions: EU, JA, US, and US delayed. |
| Autopilot for Everyone | Gemini-2.0-flash-001 (Google Gemini) | EU, US |
|
Autopilot for Everyone stores customer data in Orchestrator tenant region. Autopilot for Everyone backend service (data transit only) is deployed for the following regions: EU, JA, US, and US delayed. |
|
Autopilot for Everyone (for versions 2025.4.1 or older) | Claude-3-5-sonnet-20240620-v1:0 (Anthropic) | AU, CA, EU, IN, JA, SI, UK, US |
|
Autopilot for Everyone stores customer data in Orchestrator tenant region. Autopilot for Everyone backend service (data transit only) is deployed for the following regions: EU, JA, US, and US delayed. |
|
Autopilot for Everyone (for versions 2025.7.1 or newer) | Claude-3-7-sonnet-20250219-v1:0 (Anthropic) | AU, CA, EU, IN, SI, UK, US |
|
Autopilot for Everyone stores customer data in Orchestrator tenant region. Autopilot for Everyone backend service (data transit only) is deployed for the following regions: EU, JA, US, and US delayed. |
Autopilot for Everyone integrations:
|
GPT-4o-2024-11-20 (Azure OpenAI) | AU, CA, EU, IN, JA, SI, UK, US |
|
Only the Autopilot Teams, Slack, and Microsoft Copilot plug-ins use GPT-4o as their model. Autopilot for Everyone stores customer data in Orchestrator tenant region. Autopilot for Everyone backend service (data transit only) is deployed for the following regions: EU, JA, US, and US delayed. |
| Context Grounding | Gemini | EU, US | Routed to the closest supported region. | N/A |
| Embeddings | EU, US, JA, CA, AU, SI, IN, UK | Routed to the closest supported region. | N/A | |
| Data Fabric | Azure OpenAI | EU, JA, US | For AU, CA, IN, SI, UK regions, data is re-routed to EU. | N/A |
| Document Understanding | Azure OpenAI | AU, CA, EU, EU (delayed), IN, JA, UK, CH, US, US (delayed) |
| The generative model is deployed the same way as our specialized pre-trained models, hence, the same data privacy applies. |
| GenAI Activities | GPT 4o-x | EU, JA, US |
| N/A |
| GenAI Activities | GPT 4.1x | AU, CA, EU, IN, JA, SI, UK, US | Routed to the closest supported region. | N/A |
| GenAI Activities | GPT 5 and GPT 5 mini (Preview) | EU, US |
| N/A |
| GenAI Activities | Claude 3.5/3.7 |
|
| N/A |
| GenAI Activities | Gemini 2.5 Flash, Gemini 2.5 Pro | EU, US | Routed to the closest supported region. | N/A |
| Healing Agent | Gemini 2.5 Flash | EU, US, JA |
| N/A |
| IXP - Communications Mining | Azure OpenAI | AU, CA, EU, EU (delayed), US, US (delayed) | For SI, data is re-routed to EU. | N/A |
| IXP - Unstructured and complex documents | Azure OpenAI | AU, CA, EU, JP, US, US (delayed) | N/A | N/A |
| Gemini | AU, CA, EU, JP, US | N/A | N/A | |
| ScreenPlay | Anthropic
| EU, US |
| N/A |
Google
| EU, US, JA |
| ||
OpenAI
| EU, US |
| N/A | |
| EU |
| All other regions re-route to EU. | |
| Studio Web | Azure OpenAI | AU, EU, EU (delayed), US, US (delayed) | For the CA, IN, UK, SI regions, data is re-routed to EU. | N/A |
| Task Mining | Azure OpenAI (ChatGPT-4o) | AU, CA, EU, IN, JA, SI, UK, US | For SI, data is re-routed to the EU. | N/A |
| Test Manager | Azure OpenAI | EU, US, JA |
| N/A |
| Test Manager | Google Gemini | EU, US |
| N/A |
| Test Manager | Anthropic Claude | EU, US |
| N/A |