Share via


Features with limited regional availability

This article lists features where there is regional differentiation in feature availability.

Features that are excluded from at least one region include:

System tables

Serverless

AI and machine learning

Model serving

Note

Some features are not included in the tables below. These features, known as Designated Services, use Geographies to manage data residency when processing customer content. A Geo is a group of data center regions that Databricks uses to provide predictability and transparency regarding where your data is processed. See Databricks Designated Services.

The following tables list the Azure regions supported by Databricks. Some features are supported only in a subset of regions, and the table indicates whether or not a region supports each of these features. If a feature is supported in all regions, it is not included in the table.

Note

The Azure Government regions usgovaz and usgovva do not support any of the following features. Additionally, these regions do not support Databricks SQL or Unity Catalog.

System tables availability

Region Location System tables
australiacentral AustraliaCentral ✓
australiacentral2 AustraliaCentral2 ✓
australiaeast AustraliaEast ✓
australiasoutheast AustraliaSoutheast ✓
brazilsouth BrazilSouth ✓
canadacentral CanadaCentral ✓
canadaeast CanadaEast ✓
centralindia CentralIndia ✓
centralus CentralUS ✓
chinaeast2 ChinaEast2
chinaeast3 ChinaEast3
chinanorth2 ChinaNorth2
chinanorth3 ChinaNorth3
eastasia EastAsia ✓
eastus EastUS ✓
eastus2 EastUS2 ✓
francecentral FranceCentral ✓
germanywestcentral GermanyWestCentral ✓
japaneast JapanEast ✓
japanwest JapanWest ✓
koreacentral KoreaCentral ✓
mexicocentral MexicoCentral ✓
northcentralus NorthCentralUS ✓
northeurope NorthEurope ✓
norwayeast NorwayEast ✓
qatarcentral QatarCentral ✓
southafricanorth SouthAfricaNorth ✓
southcentralus SouthCentralUS ✓
southeastasia SoutheastAsia ✓
southindia SouthIndia ✓
swedencentral SwedenCentral ✓
switzerlandnorth SwitzerlandNorth ✓
switzerlandwest SwitzerlandWest
uaenorth UAENorth ✓
uksouth UKSouth ✓
ukwest UKWest ✓
westcentralus WestCentralUS ✓
westeurope WestEurope ✓
westindia WestIndia
westus WestUS ✓
westus2 WestUS2 ✓
westus3 WestUS3 ✓

Serverless availability

Region Location Serverless SQL warehouses Serverless compute for notebooks and jobs Mosaic AI Model Training - forecasting Serverless Lakeflow Declarative Pipelines Private connectivity Databricks Apps Managed connectors in Lakeflow Connect
australiacentral AustraliaCentral
australiacentral2 AustraliaCentral2
australiaeast AustraliaEast ✓ ✓ ✓ ✓ ✓ ✓ ✓
australiasoutheast AustraliaSoutheast
brazilsouth BrazilSouth ✓ ✓ ✓ ✓ ✓ ✓
canadacentral CanadaCentral ✓ ✓ ✓ ✓ ✓ ✓ ✓
canadaeast CanadaEast
centralindia CentralIndia ✓ ✓ ✓ ✓ ✓ ✓
centralus CentralUS ✓ ✓ ✓ ✓ ✓ ✓ ✓
chinaeast2 ChinaEast2
chinaeast3 ChinaEast3
chinanorth2 ChinaNorth2
chinanorth3 ChinaNorth3
eastasia EastAsia ✓ ✓ ✓ ✓ ✓
eastus EastUS ✓ ✓ ✓ ✓ ✓ ✓ ✓
eastus2 EastUS2 ✓ ✓ ✓ ✓ ✓ ✓ ✓
francecentral FranceCentral ✓ ✓ ✓ ✓ ✓ ✓
germanywestcentral GermanyWestCentral ✓ ✓ ✓ ✓ ✓ ✓
japaneast JapanEast ✓ ✓ ✓ ✓ ✓ ✓
japanwest JapanWest
koreacentral KoreaCentral ✓ ✓ ✓ ✓ ✓
mexicocentral MexicoCentral
northcentralus NorthCentralUS ✓ ✓ ✓ ✓ ✓ ✓ ✓
northeurope NorthEurope ✓ ✓ ✓ ✓ ✓ ✓ ✓
norwayeast NorwayEast
qatarcentral QatarCentral
southafricanorth SouthAfricaNorth
southcentralus SouthCentralUS ✓ ✓ ✓ ✓ ✓ ✓ ✓
southeastasia SoutheastAsia ✓ ✓ ✓ ✓ ✓ ✓ ✓
southindia SouthIndia
swedencentral SwedenCentral ✓ ✓ ✓ ✓ ✓ ✓
switzerlandnorth SwitzerlandNorth ✓ ✓ ✓ ✓ ✓
switzerlandwest SwitzerlandWest
uaenorth UAENorth ✓ ✓ ✓ ✓ ✓ ✓
uksouth UKSouth ✓ ✓ ✓ ✓ ✓ ✓ ✓
ukwest UKWest
westcentralus WestCentralUS ✓ ✓ ✓ ✓ ✓
westeurope WestEurope ✓ ✓ ✓ ✓ ✓ ✓ ✓
westindia WestIndia
westus WestUS ✓ ✓ ✓ ✓ ✓ ✓ ✓
westus2 WestUS2 ✓ ✓ ✓ ✓ ✓ ✓ ✓
westus3 WestUS3 ✓ ✓ ✓ ✓ ✓

AI and machine learning

Region Location Vector Search Predictive Optimization Mosaic AI Framework and Evaluation Foundation Model Fine-tuning AI Functions optimized for batch inference Lakehouse monitoring
australiacentral AustraliaCentral
australiacentral2 AustraliaCentral2
australiaeast AustraliaEast ✓ ✓ ✓   ✓† ✓
australiasoutheast AustraliaSoutheast
brazilsouth BrazilSouth ✓ ✓ ✓   ✓† ✓
canadacentral CanadaCentral ✓ ✓ ✓   ✓† ✓
canadaeast CanadaEast
centralindia CentralIndia ✓ ✓ ✓   ✓† ✓
centralus CentralUS ✓ ✓ ✓ ✓ ✓ ✓
chinaeast2 ChinaEast2
chinaeast3 ChinaEast3
chinanorth2 ChinaNorth2
chinanorth3 ChinaNorth3
eastasia EastAsia ✓   ✓   ✓† ✓
eastus EastUS ✓ ✓ ✓ ✓ ✓ ✓
eastus2 EastUS2 ✓ ✓ ✓ ✓ ✓ ✓
francecentral FranceCentral ✓ ✓ ✓ ✓ ✓
germanywestcentral GermanyWestCentral ✓ ✓ ✓ ✓ ✓
japaneast JapanEast ✓ ✓ ✓ ✓† ✓
japanwest JapanWest
koreacentral KoreaCentral ✓   ✓   ✓† ✓
mexicocentral MexicoCentral
northcentralus NorthCentralUS ✓ ✓ ✓ ✓ ✓ ✓
northeurope NorthEurope ✓ ✓ ✓ ✓ ✓
norwayeast NorwayEast
qatarcentral QatarCentral
southafricanorth SouthAfricaNorth
southcentralus SouthCentralUS ✓ ✓ ✓ ✓ ✓
southeastasia SoutheastAsia ✓ ✓ ✓   ✓† ✓
southindia SouthIndia
swedencentral SwedenCentral ✓ ✓ ✓ ✓ ✓
switzerlandnorth SwitzerlandNorth ✓ ✓ ✓ ✓
switzerlandwest SwitzerlandWest
uaenorth UAENorth ✓ ✓ ✓   ✓† ✓
uksouth UKSouth ✓ ✓ ✓ ✓ ✓† ✓
ukwest UKWest ✓ ✓
westcentralus WestCentralUS ✓
westeurope WestEurope ✓ ✓ ✓ ✓ ✓
westindia WestIndia
westus WestUS ✓ ✓ ✓ ✓ ✓ ✓
westus2 WestUS2 ✓ ✓ ✓ ✓
westus3 WestUS3 ✓ ✓ ✓ ✓

† Region supported only when cross geography routing is enabled. Foundation Model API workloads that rely on GPU compute availability, such as workloads that use models from the Meta Llama family, require cross geography routing.

Model serving regional availability

The following table summarizes region availability for Mosaic AI Model Serving capabilities.

Important

For Foundation Model APIs pay-per-token:

  • Meta Llama 4 Maverick, GTE Large (En), and Meta Llama 3.3 70B Instruct models are available in pay-per-token EU and US supported regions. See Foundation models hosted on Databricks for which models are supported in the available pay-per-token regions.
  • If your workspace is in a Model Serving region, but not in a US or EU region, your workload is routed to the US Databricks Geo. For how to enable cross region data processing, see Enable cross-Geo processing. See Databricks Designated Services for which geographies process pay-per-token workloads.

For Foundation Model APIs provisioned throughput, Meta Llama 3.1 and 3.2 models can be downloaded from HuggingFace and served using provisioned throughput endpoints. See Foundation models hosted on Databricks for supported regions and model variants.

Region Location Core Model Serving capability ** External models Foundation Model APIs (provisioned throughput) ** Foundation Model APIs (pay-per-token)
australiacentral AustraliaCentral
australiacentral2 AustraliaCentral2
australiaeast AustraliaEast ✓ ✓ ✓ ✓†
australiasoutheast AustraliaSoutheast
brazilsouth BrazilSouth ✓ ✓ ✓ ✓†
canadacentral CanadaCentral ✓ ✓ ✓ ✓†
canadaeast CanadaEast
centralindia CentralIndia ✓ ✓ ✓ ✓†
centralus CentralUS ✓ ✓ ✓ ✓
chinaeast2 ChinaEast2
chinaeast3 ChinaEast3
chinanorth2 ChinaNorth2
chinanorth3 ChinaNorth3
eastasia EastAsia ✓ ✓ ✓† ✓†
eastus EastUS ✓ ✓ ✓ ✓
eastus2 EastUS2 ✓ ✓ ✓ ✓
francecentral FranceCentral ✓ ✓ ✓ ✓
germanywestcentral GermanyWestCentral ✓ ✓ ✓ ✓
japaneast JapanEast ✓ ✓ ✓† ✓†
japanwest JapanWest
koreacentral KoreaCentral ✓ ✓ ✓† ✓†
mexicocentral MexicoCentral
northcentralus NorthCentralUS ✓ ✓ ✓ ✓
northeurope NorthEurope ✓ ✓ ✓ ✓
norwayeast NorwayEast
qatarcentral QatarCentral
southafricanorth SouthAfricaNorth
southcentralus SouthCentralUS ✓ ✓ ✓ ✓
southeastasia SoutheastAsia ✓ ✓ ✓† ✓†
southindia SouthIndia
swedencentral SwedenCentral ✓ ✓ ✓ ✓
switzerlandnorth SwitzerlandNorth ✓ ✓ ✓ ✓
switzerlandwest SwitzerlandWest
uaenorth UAENorth ✓ ✓ ✓† ✓†
uksouth UKSouth ✓ ✓ ✓† ✓†
ukwest UKWest
westcentralus WestCentralUS
westeurope WestEurope ✓ ✓ ✓ ✓
westindia WestIndia
westus WestUS ✓ ✓ ✓ ✓
westus2 WestUS2 ✓ (preview) ✓ (preview) ✓ ✓ (preview)
westus3 WestUS3 ✓ ✓ ✓ ✓

** Includes GPU support

† Region supported only when cross geography routing is enabled. Foundation Model API workloads that rely on GPU compute availability, such as workloads that use models from the Meta Llama family, require cross geography routing.