Google - Amazon Bedrock
Services or capabilities described in AWS documentation might vary by Region. To see the differences applicable to the AWS European Sovereign Cloud Region, see the AWS European Sovereign Cloud User Guide.

Google

The following Google models are available in Amazon Bedrock:

Model Description
Gemma 4 31BGemma 4 31B is Google's 30.7-billion parameter dense model with built-in reasoning, native function calling, and multimodal input across text and image, supporting a 256K token context window.
Gemma 4 26B-A4BGemma 4 26B-A4B is Google's mixture-of-experts model with 25.2 billion total parameters and 3.8 billion active per token, delivering cost-efficient inference with built-in reasoning, native function calling, and multimodal input across text and image, supporting a 256K token context window.
Gemma 4 E2BGemma 4 E2B is Google's compact model with 5.1 billion total parameters and 2.3 billion effective parameters, designed for low-latency workloads with built-in reasoning, native function calling, and multimodal input across text and image, supporting a 128K token context window.
Gemma 3 12B ITGemma 3 12B IT is Google's 12-billion parameter open model with instruction tuning, supporting text and image inputs with a 128K context window.
Gemma 3 27B PTGemma 3 27B PT is Google's 27-billion parameter open model, the largest in the Gemma 3 family, with multimodal capabilities and a 128K context window.
Gemma 3 4B ITGemma 3 4B IT is Google's compact 4-billion parameter open model with instruction tuning, designed for on-device and edge deployment.