Introducing Gemma 4 models on Amazon Bedrock

The Gemma 4 family of open-weight models is now available on Amazon Bedrock, offering a range of instruction-tuned variants with dense and mixture-of-experts architectures. The models, built by Google DeepMind, provide built-in reasoning, native function calling, and multimodal input across text and image, with a focus on intelligence-per-parameter. With Amazon Bedrock, organizations can access leading open-weight foundation models without compromising on data protection, regulatory alignment, or operational control. The Gemma 4 family includes three variants: Gemma 4 31B, Gemma 4 26B-A4B, and Gemma 4 E2B, which can be used to build multimodal agents, lightweight applications, and document understanding pipelines.

⚡ Key Takeaways

• The Gemma 4 family includes three instruction-tuned variants: Gemma 4 31B, Gemma 4 26B-A4B, and Gemma 4 E2B, with 30.7B, 25.2B, and 5.1B total parameters, respectively.
• The models support built-in reasoning, native function calling, and multimodal input across text and image, with out-of-the-box support for over 35 languages.
• Amazon Bedrock provides a fully managed service for accessing the Gemma 4 models, with inference running entirely on infrastructure operated by AWS and security and privacy controls.
• The models can be accessed through a fully managed AWS service without provisioning infrastructure, hosting model weights, or operating inference stacks.
• The Gemma 4 family has a context window of 256K tokens for Gemma 4 31B and Gemma 4 26B-A4B, and 128K tokens for Gemma 4 E2B.

The availability of the Gemma 4 family on Amazon Bedrock provides organizations with a range of open-weight models that can be used for various applications, including multimodal agents, lightweight applications, and document understanding pipelines, while ensuring data protection, regulatory alignment, and operational control. This can help engineers building AI systems to develop more intelligen

✅ Practical Steps

1. Access the Gemma 4 models through the Amazon Bedrock model catalog.
2. Use the Amazon Bedrock APIs to integrate the Gemma 4 models into your applications.
3. Evaluate the model architecture and training methodology, and fine-tune the models on proprietary data when customization is required.

Want the full story? Read the original article.