Regional availability of services and features

IBM watsonx is deployed on a multi-zone region network. The availability of services and features can vary across regional data centers.

Regional availability of services on IBM Cloud

To access the list of available services and view regional availability of each service on IBM Cloud, see the IBM watsonx services catalog.

IBM watsonx is available in the following regional data centers:

  • Dallas (us-south), in Texas, US
  • Frankfurt (eu-de), in Germany
  • London (eu-gb), in England
  • Tokyo (jp-tok), in Japan
  • Sydney (au-syd), in Australia
  • Toronto (ca-tor), in Canada

Regional availability of watsonx services on IBM Cloud

You can access the watsonx.ai Studio, watsonx.ai Runtime, and watsonx.governance services from all regional data centers.

Note:

The watsonx.ai Runtime service has some restrictions when you provision the service instance with the HIPAA-Ready plan. See HIPAA-Ready plan details.

Regional availability of the Cloud Object Storage service on IBM Cloud

The region for the Cloud Object Storage service is Global. Cloud Object Storage buckets for workspaces are Regional buckets. For more information, see IBM Cloud docs: Cloud Object Storage endpoints and storage locations.

Regional availability of the watsonx services on AWS

The IBM watsonx.ai service is available in the following regional data centers:

  • Mumbai (ap-south-1), in India

The watsonx.governance service is available in the following regional data centers:

  • Mumbai (ap-south-1), in India (Provides model management and Governance Console.)
  • Northern Virginia (us-east-1), in US (Provides Governance Console.)
  • Frankfurt (eu-central-1), in Germany (Provides Governance Console.)

Regional availability of watsonx features

Some features and tools for the watsonx.ai Studio, watsonx.ai Runtime, and watsonx.governance services are not available in every region.

The following table lists the regional availability of features that are not available in all regions. Features that are available in all regions are not listed.

Regional availability of watsonx features
Feature Platform/Service Dallas Frankfurt London Mumbai (AWS) Sydney Tokyo Toronto
IBM Orchestration Pipelines watsonx.ai Studio ✓ ✓ ✓ ✓
RStudio watsonx.ai Studio ✓ ✓ ✓ ✓
Use multibyte characters in user-input fields in Data Refinery. watsonx.ai Studio ✓ ✓ ✓ ✓ ✓
Custom foundation models watsonx.ai Runtime ✓ ✓ ✓ ✓ ✓
Deploy on demand foundation models watsonx.ai Runtime ✓ ✓ ✓ ✓ ✓
Text classification API watsonx.ai Runtime ✓ ✓ ✓ ✓ ✓ ✓
Foundation model fine tuning watsonx.ai Runtime ✓ ✓ ✓ ✓
Evaluation Studio watsonx.governance ✓ ✓ ✓ ✓ ✓ ✓ ✓
Tuned model evaluations watsonx.governance ✓ ✓ ✓ ✓ ✓ ✓
Only one instance per region watsonx.governance ✓ ✓ ✓ ✓ ✓ ✓ ✓
Integration with Governance console watsonx.governance ✓ ✓ ✓ ✓
Governed agentic catalog watsonx.governance ✓ ✓ ✓ ✓ ✓ ✓
Monitoring watsonx Orchestrate agents watsonx.governance ✓ ✓ ✓ ✓ ✓ ✓

Regional compliance with information security and privacy standards

Depending on the data center that you work with, different services might comply with different information security and privacy standards. For details, see the Regional compliance with information security and privacy standards table.

Regional compliance with information security and privacy standards
Compliance Program Service Dallas Frankfurt London Mumbai (AWS) Sydney Tokyo Toronto
ISO/IEC 27001 watsonx.ai Studio (Formerly known as Watson Studio) ✓ ✓ ✓ ✓ ✓ ✓
ISO/IEC 27001 watsonx.ai Runtime (Formerly known as Watson Machine Learning) ✓ ✓ ✓ ✓ ✓ ✓
ISO/IEC 27001 watsonx.ai ✓ ✓ ✓ ✓ ✓ ✓
ISO/IEC 27001 watsonx.governance ✓
ISO/IEC 27017 watsonx.ai Studio (Formerly known as Watson Studio) ✓ ✓ ✓ ✓ ✓ ✓
ISO/IEC 27017 watsonx.ai Runtime (Formerly known as Watson Machine Learning) ✓ ✓ ✓ ✓ ✓ ✓
ISO/IEC 27017 watsonx.ai ✓ ✓ ✓ ✓ ✓ ✓
ISO/IEC 27017 watsonx.governance ✓
ISO/IEC 27018 watsonx.ai Studio (Formerly known as Watson Studio) ✓ ✓ ✓ ✓ ✓ ✓
ISO/IEC 27018 watsonx.ai Runtime (Formerly known as Watson Machine Learning) ✓ ✓ ✓ ✓ ✓ ✓
ISO/IEC 27018 watsonx.ai ✓ ✓ ✓ ✓ ✓ ✓
ISO/IEC 27018 watsonx.governance ✓
SOC1 Type2 watsonx.ai Studio (Formerly known as Watson Studio) ✓ ✓ ✓ ✓
SOC1 Type2 watsonx.ai Runtime (Formerly known as Watson Machine Learning) ✓ ✓ ✓ ✓
SOC1 Type2 watsonx.ai ✓ ✓ ✓ ✓
SOC2 Type2 watsonx.ai Studio (Formerly known as Watson Studio) ✓ ✓ ✓ ✓
SOC2 Type2 watsonx.ai Runtime (Formerly known as Watson Machine Learning) ✓ ✓ ✓ ✓
SOC2 Type2 watsonx.ai ✓ ✓ ✓ ✓

Regional availability of foundation models

The regional availability of foundation models varies based on how the model is deployed in watsonx.ai, the model type, and the tasks you perform with the foundation model.

Based on your use case, see:

For more details about the models, such as pricing and API model IDs, see the following topics:

Regional availability of provided foundation models that are ready to use

The following table shows the data centers where you can inference multitenant foundation models that are provided by IBM. A checkmark indicates that the model is hosted in the region.

Data center support for inferencing foundation models provided by IBM
Model name Dallas Frankfurt London Mumbai (AWS) Sydney Tokyo Toronto
allam-1-13b-instruct ✓
gpt-oss-120b ✓
granite-4-h-small ✓ ✓
granite-3-3-8b-instruct ✓ ✓ ✓
granite-3-2-8b-instruct ✓ ✓
granite-3-8b-instruct ✓ ✓ ✓ ✓ ✓ ✓
granite-guardian-3-8b ✓
granite-8b-code-instruct ✓ ✓ ✓
granite-ttm-512-96-r2 ✓ ✓ ✓ ✓ ✓ ✓
granite-ttm-512-1024-r2 ✓ ✓ ✓ ✓ ✓ ✓
granite-ttm-512-1536-r2 ✓ ✓ ✓ ✓ ✓ ✓
llama-4-maverick-17b-128e-instruct-fp8 ✓ ✓ ✓ ✓ ✓
llama-3-3-70b-instruct ✓ ✓ ✓ ✓ ✓ ✓
llama-3-2-11b-vision-instruct ✓ ✓ ✓ ✓ ✓
llama-3-2-90b-vision-instruct ✓ ✓ ✓ ✓
llama-guard-3-11b-vision ✓ ✓
llama-3-405b-instruct ✓
mistral-large-2512 ✓
mistral-medium-2505 ✓ ✓ ✓
mistral-small-3-1-24b-instruct-2503 ✓ ✓ ✓ ✓ ✓ ✓

Regional availability of deploy on demand foundation models

The following table shows the data centers where you can deploy foundation models on-demand for the dedicated use of your organization. A checkmark indicates that the model is hosted in the region.

Data center support for deploy on demand foundation models provided by IBM
Model name Dallas Frankfurt London Mumbai (AWS) Sydney Tokyo Toronto
granite-4-h-small ✓ ✓ ✓ ✓ ✓
granite-4-h-tiny ✓ ✓ ✓ ✓ ✓
granite-4-h-micro ✓ ✓ ✓ ✓ ✓
granite-vision-3-3-2b ✓ ✓ ✓ ✓
granite-3-1-8b-base ✓ ✓ ✓ ✓ ✓
granite-3-3-8b-instruct ✓ ✓ ✓ ✓ ✓
granite-3-3-2b-instruct ✓ ✓ ✓ ✓ ✓
granite-3-8b-base ✓ ✓ ✓ ✓ ✓
granite-7b-lab ✓ ✓ ✓ ✓ ✓
granite-8b-japanese ✓ ✓ ✓ ✓ ✓
granite-20b-multilingual ✓ ✓ ✓ ✓ ✓
granite-13b-chat-v2 ✓ ✓ ✓ ✓ ✓
granite-3b-code-instruct ✓ ✓ ✓ ✓ ✓
granite-8b-code-instruct ✓ ✓ ✓ ✓ ✓
granite-20b-code-instruct ✓ ✓ ✓ ✓ ✓
granite-34b-code-instruct ✓ ✓ ✓ ✓ ✓
granite-20b-code-base-schema-linking ✓ ✓ ✓ ✓ ✓
granite-20b-code-base-sql-gen ✓ ✓ ✓ ✓ ✓
allam-1-13b-instruct ✓ ✓ ✓ ✓ ✓
codestral-2501 ✓ ✓ ✓ ✓ ✓
codellama-34b-instruct-hf ✓ ✓ ✓ ✓ ✓
deepseek-r1-distill-llama-8b ✓ ✓ ✓ ✓ ✓
deepseek-r1-distill-llama-70b ✓ ✓ ✓ ✓ ✓
eurollm-1-7b-instruct ✓ ✓ ✓ ✓ ✓
eurollm-9b-instruct ✓ ✓ ✓ ✓ ✓
gpt-oss-20b ✓ ✓ ✓ ✓ ✓
gpt-oss-120b ✓ ✓ ✓ ✓ ✓
llama-2-13b-chat ✓ ✓ ✓ ✓ ✓
llama-2-70b-chat ✓ ✓ ✓ ✓ ✓
llama-3-8b-instruct ✓ ✓ ✓ ✓ ✓
llama-3-70b-instruct ✓ ✓ ✓ ✓ ✓
llama-3-1-8b ✓ ✓ ✓ ✓ ✓
llama-3-1-8b-instruct ✓ ✓ ✓ ✓ ✓
llama-3-1-70b ✓ ✓ ✓ ✓ ✓
llama-3-1-70b-instruct ✓ ✓ ✓ ✓ ✓
llama-3-1-70b-gptq ✓ ✓ ✓ ✓ ✓
llama-3-2-11b-vision-instruct ✓ ✓ ✓ ✓ ✓
llama-3-3-70b-instruct ✓ ✓ ✓ ✓ ✓
llama-3-3-70b-instruct-hf ✓ ✓ ✓ ✓ ✓
llama-4-maverick-17b-128e-instruct-fp8 ✓ ✓ ✓ ✓
llama-4-maverick-17b-128e-instruct-int4 ✓ ✓ ✓ ✓
ministral-8b-instruct-2410 ✓ ✓ ✓ ✓ ✓
mistral-large-instruct-2407 ✓ ✓ ✓ ✓
mistral-large-instruct-2411 ✓ ✓ ✓ ✓
mistral-medium-2505 ✓ ✓ ✓ ✓
mistral-medium-2508 ✓ ✓ ✓ ✓
mistral-nemo-instruct-2407 ✓ ✓ ✓ ✓ ✓
mixtral-8x7b-base ✓ ✓ ✓ ✓ ✓
mixtral-8x7b-instruct-v01 ✓ ✓ ✓ ✓ ✓
mt0-xxl-13b ✓ ✓ ✓ ✓ ✓
poro-34b-chat ✓ ✓ ✓ ✓ ✓

Regional availability of foundation model tuning

The following table shows the data centers where tuning a foundation model is supported and the models that can be tuned.

A checkmark indicates that you can use the Tuning Studio to tune the foundation model in the region.

Data center support for foundation model fine tuning with LoRA
Model name Dallas Frankfurt London Mumbai (AWS) Sydney Tokyo Toronto
granite-3-1-8b-base ✓ ✓
llama-3-1-8b ✓ ✓

Regional availability of encoder-only models

The following table shows the data centers where embedding models and reranker models are available. A checkmark indicates that you can use the model in the region.

Data center support for encoder-only models
Model Dallas Frankfurt London Mumbai (AWS) Sydney Tokyo Toronto
all-minilm-l6-v2 ✓ ✓ ✓ ✓
granite-embedding-278m-multilingual ✓ ✓ ✓ ✓ ✓ ✓
ms-marco-minilm-l-12-v2 ✓ ✓ ✓ ✓ ✓ ✓ ✓
multilingual-e5-large ✓ ✓ ✓ ✓ ✓ ✓ ✓
slate-125m-english.rtrvr-v2 ✓ ✓ ✓ ✓ ✓ ✓ ✓
slate-30m-english.rtrvr-v2 ✓ ✓ ✓ ✓ ✓ ✓ ✓

Learn more