Skip to main content

Model availability by region

Check the tables for model and model-family support in each region and serving option. To browse models by provider, see available models in Unity Gateway.

The tables describe model support. To query a model through Unity Gateway, check for the corresponding model service in your workspace. Availability depends on region, cross-Geo settings, and available serving capacity. Required permissions still apply.

Foundation models hosted on Databricks​

Databricks hosts state-of-the-art open foundation models. These models are made available using Foundation Model APIs. To restrict which Databricks-hosted foundation models your organization can use, see foundation model Unity Catalog permissions.

The following table summarizes which Databricks-hosted models and model families are supported in each region based on the Model Serving feature.

important
  • Google Gemini 3 Flash is hosted on global endpoints and requires cross geography routing to be enabled for every region.

  • Meta Llama 4 Maverick is available for Foundation Model APIs provisioned throughput workloads in Public Preview.

  • Google Gemini 2.5 Flash and Pro are deprecated and are planned for retirement on October 2, 2026. See Deprecated and retired models for the recommended replacement model and guidance for how to migrate during deprecation.

  • Anthropic Claude Sonnet 4 is deprecated and planned for retirement on October 9, 2026. See Deprecated and retired models for its retirement date, recommended replacement model and guidance for how to migrate during deprecation.

  • Anthropic Claude 3.7 Sonnet is no longer available. See Deprecated and retired models for the recommended replacement model and guidance for how to migrate during deprecation.

  • Several older model families have been retired. See Deprecated and retired models for the full list of retired models and recommended replacements.

Region

Foundation Model APIs pay-per-token

AI Functions (batch inference)

Foundation Model APIs provisioned throughput

ap-northeast-1

The following models are supported:

The following models are supported:

  • The following model families are supported for real-time inference:
    • OpenAI GPT OSS 120B
    • OpenAI GPT OSS 20B
    • Google Gemma 3 12B
    • Alibaba Cloud Qwen3.5 122B A10B (preview)
    • Meta Llama 4 Maverick⥂ (preview)
    • Meta Llama 3.3
    • Meta Llama 3.2 3B
    • Meta Llama 3.2 1B
    • Meta Llama 3.1
    • GTE v1.5 (English)
    • BGE v1.5 (English)

ap-northeast-2

The following models are supported:

The following models are supported:

  • The following model families are supported for real-time inference:
    • OpenAI GPT OSS 120B
    • OpenAI GPT OSS 20B
    • Google Gemma 3 12B
    • Meta Llama 4 Maverick⥂ (preview)
    • Meta Llama 3.3
    • Meta Llama 3.2 3B
    • Meta Llama 3.2 1B
    • Meta Llama 3.1
    • GTE v1.5 (English)
    • BGE v1.5 (English)

ap-south-1

The following models are supported:

The following models are supported:

  • The following model families are supported for real-time inference:
    • OpenAI GPT OSS 120B
    • OpenAI GPT OSS 20B⥂
    • Google Gemma 3 12B
    • Alibaba Cloud Qwen3.5 122B A10B (preview)
    • Meta Llama 4 Maverick⥂ (preview)
    • Meta Llama 3.3
    • Meta Llama 3.2 3B⥂
    • Meta Llama 3.2 1B⥂
    • Meta Llama 3.1
    • GTE v1.5 (English)
    • BGE v1.5 (English)

ap-southeast-1

The following models are supported:

The following models are supported:

  • The following model families are supported and optimized for batch inference:

    • OpenAI GPT OSS 120B
    • OpenAI GPT OSS 20B
    • Google Gemma 3 12B⥂
    • Meta Llama 4 Maverick⥂ (preview)
  • The following model families are supported for real-time inference:

    • OpenAI GPT OSS 120B
    • OpenAI GPT OSS 20B
    • Google Gemma 3 12B⥂
    • Meta Llama 4 Maverick⥂ (preview)
    • Meta Llama 3.3
    • Meta Llama 3.2 3B
    • Meta Llama 3.2 1B
    • Meta Llama 3.1
    • GTE v1.5 (English)
    • BGE v1.5 (English)

ap-southeast-2

The following models are supported:

The following models are supported:

  • The following model families are supported and optimized for batch inference:

    • OpenAI GPT OSS 120B⥂
    • OpenAI GPT OSS 20B⥂
    • Google Gemma 3 12B⥂
    • Meta Llama 4 Maverick⥂ (preview)
  • The following model families are supported for real-time inference:

    • OpenAI GPT OSS 120B
    • OpenAI GPT OSS 20B
    • Google Gemma 3 12B
    • Meta Llama 4 Maverick⥂ (preview)
    • Meta Llama 3.3
    • Meta Llama 3.2 3B
    • Meta Llama 3.2 1B
    • Meta Llama 3.1
    • GTE v1.5 (English)
    • BGE v1.5 (English)

ca-central-1

The following models are supported:

The following models are supported:

  • The following model families are supported and optimized for batch inference:

    • OpenAI GPT OSS 120B⥂
    • OpenAI GPT OSS 20B⥂
    • Google Gemma 3 12B
    • Meta Llama 4 Maverick (preview)
  • The following model families are supported for real-time inference:

    • OpenAI GPT OSS 120B⥂
    • OpenAI GPT OSS 20B⥂
    • Google Gemma 3 12B
    • Meta Llama 4 Maverick (preview)
    • Meta Llama 3.3
    • Meta Llama 3.2 3B
    • Meta Llama 3.2 1B
    • Meta Llama 3.1
    • GTE v1.5 (English)
    • BGE v1.5 (English)

eu-central-1

The following models are supported:

The following models are supported:

  • The following model families are supported and optimized for batch inference:

    • OpenAI GPT OSS 120B
    • OpenAI GPT OSS 20B
    • Google Gemma 3 12B
    • Meta Llama 4 Maverick (preview)
  • The following model families are supported for real-time inference:

    • OpenAI GPT OSS 120B
    • OpenAI GPT OSS 20B
    • Google Gemma 3 12B
    • Alibaba Cloud Qwen3.5 122B A10B (preview)
    • Meta Llama 4 Maverick (preview)
    • Meta Llama 3.3
    • Meta Llama 3.2 3B
    • Meta Llama 3.2 1B
    • Meta Llama 3.1
    • GTE v1.5 (English)
    • BGE v1.5 (English)

eu-west-1

The following models are supported:

The following models are supported:

  • The following model families are supported and optimized for batch inference:

    • OpenAI GPT OSS 120B
    • OpenAI GPT OSS 20B
    • Google Gemma 3 12B
    • Meta Llama 4 Maverick (preview)
  • The following model families are supported for real-time inference:

    • OpenAI GPT OSS 120B
    • OpenAI GPT OSS 20B
    • Google Gemma 3 12B
    • Alibaba Cloud Qwen3.5 122B A10B (preview)
    • Meta Llama 4 Maverick (preview)
    • Meta Llama 3.3
    • Meta Llama 3.2 3B
    • Meta Llama 3.2 1B
    • Meta Llama 3.1
    • GTE v1.5 (English)
    • BGE v1.5 (English)

eu-west-2

The following models are supported:

The following models are supported:

  • The following model families are supported and optimized for batch inference:

    • OpenAI GPT OSS 120B
    • OpenAI GPT OSS 20B
    • Google Gemma 3 12B
    • Meta Llama 4 Maverick (preview)
  • The following model families are supported for real-time inference:

    • OpenAI GPT OSS 120B
    • OpenAI GPT OSS 20B
    • Google Gemma 3 12B
    • Meta Llama 4 Maverick (preview)
    • Meta Llama 3.3
    • Meta Llama 3.2 3B
    • Meta Llama 3.2 1B
    • Meta Llama 3.1
    • GTE v1.5 (English)
    • BGE v1.5 (English)

eu-west-3

Not supported

Not supported

Not supported

sa-east-1

The following models are supported:

The following models are supported:

  • The following model families are supported and optimized for batch inference:

    • OpenAI GPT OSS 120B⥂
    • OpenAI GPT OSS 20B⥂
    • Google Gemma 3 12B
    • Meta Llama 4 Maverick (preview)
  • The following model families are supported for real-time inference:

    • OpenAI GPT OSS 120B⥂
    • OpenAI GPT OSS 20B⥂
    • Google Gemma 3 12B
    • Meta Llama 4 Maverick (preview)
    • Meta Llama 3.3
    • Meta Llama 3.2 3B
    • Meta Llama 3.2 1B
    • Meta Llama 3.1
    • GTE v1.5 (English)
    • BGE v1.5 (English)

us-east-1

The following models are supported:

The following models are supported:

The following model families are supported:

  • OpenAI GPT OSS 120B
  • OpenAI GPT OSS 20B
  • Google Gemma 3 12B
  • Meta Llama 4 Maverick (preview)
  • Meta Llama 3.3
  • Meta Llama 3.2 3B
  • Meta Llama 3.2 1B
  • Meta Llama 3.1
  • GTE v1.5 (English)
  • BGE v1.5 (English)

us-east-2

The following models are supported:

The following models are supported:

The following model families are supported for real-time inference:

  • OpenAI GPT OSS 120B
  • OpenAI GPT OSS 20B
  • Google Gemma 3 12B
  • Meta Llama 4 Maverick (preview)
  • Meta Llama 3.3
  • Meta Llama 3.2 3B
  • Meta Llama 3.2 1B
  • Meta Llama 3.1
  • GTE v1.5 (English)
  • BGE v1.5 (English)

us-gov-west-1

The following models are supported:

The following models are supported:

Not supported

us-west-1

Not supported

Not supported

Not supported

us-west-2

The following models are supported:

The following models are supported:

The following model families are supported:

  • OpenAI GPT OSS 120B
  • OpenAI GPT OSS 20B
  • Google Gemma 3 12B
  • Alibaba Cloud Qwen3.5 122B A10B (preview)
  • Meta Llama 4 Maverick (preview)
  • Meta Llama 3.3
  • Meta Llama 3.2 3B
  • Meta Llama 3.2 1B
  • Meta Llama 3.1
  • GTE v1.5 (English)
  • BGE v1.5 (English)

Region

Foundation Model APIs pay-per-token

AI Functions (batch inference)

Foundation Model APIs provisioned throughput

ap-northeast-1

The following models are supported:

The following models are supported:

  • The following model families are supported for real-time inference:
    • OpenAI GPT OSS 120B
    • OpenAI GPT OSS 20B
    • Google Gemma 3 12B
    • Alibaba Cloud Qwen3.5 122B A10B (preview)
    • Meta Llama 4 Maverick⥂ (preview)
    • Meta Llama 3.3
    • Meta Llama 3.2 3B
    • Meta Llama 3.2 1B
    • Meta Llama 3.1
    • GTE v1.5 (English)
    • BGE v1.5 (English)

ap-northeast-2

The following models are supported:

The following models are supported:

  • The following model families are supported for real-time inference:
    • OpenAI GPT OSS 120B
    • OpenAI GPT OSS 20B
    • Google Gemma 3 12B
    • Meta Llama 4 Maverick⥂ (preview)
    • Meta Llama 3.3
    • Meta Llama 3.2 3B
    • Meta Llama 3.2 1B
    • Meta Llama 3.1
    • GTE v1.5 (English)
    • BGE v1.5 (English)

ap-south-1

The following models are supported:

The following models are supported:

  • The following model families are supported for real-time inference:
    • OpenAI GPT OSS 120B
    • OpenAI GPT OSS 20B⥂
    • Google Gemma 3 12B
    • Alibaba Cloud Qwen3.5 122B A10B (preview)
    • Meta Llama 4 Maverick⥂ (preview)
    • Meta Llama 3.3
    • Meta Llama 3.2 3B⥂
    • Meta Llama 3.2 1B⥂
    • Meta Llama 3.1
    • GTE v1.5 (English)
    • BGE v1.5 (English)

ap-southeast-1

The following models are supported:

The following models are supported:

  • The following model families are supported and optimized for batch inference:

    • OpenAI GPT OSS 120B
    • OpenAI GPT OSS 20B
    • Google Gemma 3 12B⥂
    • Meta Llama 4 Maverick⥂ (preview)
  • The following model families are supported for real-time inference:

    • OpenAI GPT OSS 120B
    • OpenAI GPT OSS 20B
    • Google Gemma 3 12B⥂
    • Meta Llama 4 Maverick⥂ (preview)
    • Meta Llama 3.3
    • Meta Llama 3.2 3B
    • Meta Llama 3.2 1B
    • Meta Llama 3.1
    • GTE v1.5 (English)
    • BGE v1.5 (English)

ap-southeast-2

The following models are supported:

The following models are supported:

  • The following model families are supported and optimized for batch inference:

    • OpenAI GPT OSS 120B⥂
    • OpenAI GPT OSS 20B⥂
    • Google Gemma 3 12B⥂
    • Meta Llama 4 Maverick⥂ (preview)
  • The following model families are supported for real-time inference:

    • OpenAI GPT OSS 120B
    • OpenAI GPT OSS 20B
    • Google Gemma 3 12B
    • Meta Llama 4 Maverick⥂ (preview)
    • Meta Llama 3.3
    • Meta Llama 3.2 3B
    • Meta Llama 3.2 1B
    • Meta Llama 3.1
    • GTE v1.5 (English)
    • BGE v1.5 (English)

ca-central-1

The following models are supported:

The following models are supported:

  • The following model families are supported and optimized for batch inference:

    • OpenAI GPT OSS 120B⥂
    • OpenAI GPT OSS 20B⥂
    • Google Gemma 3 12B
    • Meta Llama 4 Maverick (preview)
  • The following model families are supported for real-time inference:

    • OpenAI GPT OSS 120B⥂
    • OpenAI GPT OSS 20B⥂
    • Google Gemma 3 12B
    • Meta Llama 4 Maverick (preview)
    • Meta Llama 3.3
    • Meta Llama 3.2 3B
    • Meta Llama 3.2 1B
    • Meta Llama 3.1
    • GTE v1.5 (English)
    • BGE v1.5 (English)

eu-central-1

The following models are supported:

The following models are supported:

  • The following model families are supported and optimized for batch inference:

    • OpenAI GPT OSS 120B
    • OpenAI GPT OSS 20B
    • Google Gemma 3 12B
    • Meta Llama 4 Maverick (preview)
  • The following model families are supported for real-time inference:

    • OpenAI GPT OSS 120B
    • OpenAI GPT OSS 20B
    • Google Gemma 3 12B
    • Alibaba Cloud Qwen3.5 122B A10B (preview)
    • Meta Llama 4 Maverick (preview)
    • Meta Llama 3.3
    • Meta Llama 3.2 3B
    • Meta Llama 3.2 1B
    • Meta Llama 3.1
    • GTE v1.5 (English)
    • BGE v1.5 (English)

eu-west-1

The following models are supported:

The following models are supported:

  • The following model families are supported and optimized for batch inference:

    • OpenAI GPT OSS 120B
    • OpenAI GPT OSS 20B
    • Google Gemma 3 12B
    • Meta Llama 4 Maverick (preview)
  • The following model families are supported for real-time inference:

    • OpenAI GPT OSS 120B
    • OpenAI GPT OSS 20B
    • Google Gemma 3 12B
    • Alibaba Cloud Qwen3.5 122B A10B (preview)
    • Meta Llama 4 Maverick (preview)
    • Meta Llama 3.3
    • Meta Llama 3.2 3B
    • Meta Llama 3.2 1B
    • Meta Llama 3.1
    • GTE v1.5 (English)
    • BGE v1.5 (English)

eu-west-2

The following models are supported:

The following models are supported:

  • The following model families are supported and optimized for batch inference:

    • OpenAI GPT OSS 120B
    • OpenAI GPT OSS 20B
    • Google Gemma 3 12B
    • Meta Llama 4 Maverick (preview)
  • The following model families are supported for real-time inference:

    • OpenAI GPT OSS 120B
    • OpenAI GPT OSS 20B
    • Google Gemma 3 12B
    • Meta Llama 4 Maverick (preview)
    • Meta Llama 3.3
    • Meta Llama 3.2 3B
    • Meta Llama 3.2 1B
    • Meta Llama 3.1
    • GTE v1.5 (English)
    • BGE v1.5 (English)

eu-west-3

Not supported

Not supported

Not supported

sa-east-1

The following models are supported:

The following models are supported:

  • The following model families are supported and optimized for batch inference:

    • OpenAI GPT OSS 120B⥂
    • OpenAI GPT OSS 20B⥂
    • Google Gemma 3 12B
    • Meta Llama 4 Maverick (preview)
  • The following model families are supported for real-time inference:

    • OpenAI GPT OSS 120B⥂
    • OpenAI GPT OSS 20B⥂
    • Google Gemma 3 12B
    • Meta Llama 4 Maverick (preview)
    • Meta Llama 3.3
    • Meta Llama 3.2 3B
    • Meta Llama 3.2 1B
    • Meta Llama 3.1
    • GTE v1.5 (English)
    • BGE v1.5 (English)

us-east-1

The following models are supported:

The following models are supported:

The following model families are supported:

  • OpenAI GPT OSS 120B
  • OpenAI GPT OSS 20B
  • Google Gemma 3 12B
  • Meta Llama 4 Maverick (preview)
  • Meta Llama 3.3
  • Meta Llama 3.2 3B
  • Meta Llama 3.2 1B
  • Meta Llama 3.1
  • GTE v1.5 (English)
  • BGE v1.5 (English)

us-east-2

The following models are supported:

The following models are supported:

The following model families are supported for real-time inference:

  • OpenAI GPT OSS 120B
  • OpenAI GPT OSS 20B
  • Google Gemma 3 12B
  • Meta Llama 4 Maverick (preview)
  • Meta Llama 3.3
  • Meta Llama 3.2 3B
  • Meta Llama 3.2 1B
  • Meta Llama 3.1
  • GTE v1.5 (English)
  • BGE v1.5 (English)

us-gov-west-1

The following models are supported:

The following models are supported:

Not supported

us-west-1

Not supported

Not supported

Not supported

us-west-2

The following models are supported:

The following models are supported:

The following model families are supported:

  • OpenAI GPT OSS 120B
  • OpenAI GPT OSS 20B
  • Google Gemma 3 12B
  • Alibaba Cloud Qwen3.5 122B A10B (preview)
  • Meta Llama 4 Maverick (preview)
  • Meta Llama 3.3
  • Meta Llama 3.2 3B
  • Meta Llama 3.2 1B
  • Meta Llama 3.1
  • GTE v1.5 (English)
  • BGE v1.5 (English)

⥂ This model is supported based on GPU availability and requires cross geography routing to be enabled.