Introducing Gemma 4 models on Amazon Bedrock

Summary

Google DeepMind's Gemma 4 models are now available on Amazon Bedrock, offering three instruction-tuned variants with dense and MoE architectures, supporting reasoning, function calling, and multimodal input.

Why it matters

The availability of Gemma 4 on Amazon Bedrock expands the options for developers and enterprises to deploy advanced AI models with flexible architectures. The inclusion of reasoning, function calling, and multimodal capabilities enables more sophisticated applications, from complex problem-solving to interactive agents.

Related tools

• Browse AI tools for products in this space
• Model library for weights and APIs
• Rankings for curated shortlists

Impact on AI tools/models

Gemma 4's integration into Amazon Bedrock signifies a trend toward making cutting-edge models accessible via managed services. This reduces the barrier for developers to experiment with state-of-the-art architectures like MoE and multimodal models without managing infrastructure.

What to watch

• AI news for updates on model releases
• Model library for detailed specs and comparisons
• Rankings for performance benchmarks

FAQ

What are Gemma 4 models? Gemma 4 models are Google DeepMind's latest instruction-tuned models with dense and MoE architectures, supporting reasoning, function calling, and multimodal input.

Where are Gemma 4 models available? They are now available on Amazon Bedrock.

What architectures do Gemma 4 models use? They use dense and mixture-of-experts (MoE) architectures.