Gemma 4 Models Now Available on Amazon Bedrock

Google DeepMind's Gemma 4 family of open-weight AI models is now accessible via Amazon Bedrock. These models, released under the Apache 2.0 license, are engineered for high intelligence relative to their parameter count, making them suitable for diverse deployment needs. The lineup features three instruction-tuned versions: Gemma 4 31B, Gemma 4 26B-A4B, and Gemma 4 E2B. They support both dense and mixture-of-experts architectures, optimizing performance by activating only necessary parameters per query. Key capabilities include advanced reasoning, native function calling, and the ability to process both text and image inputs.