ReleaseBytes
aws AWS What's New ·

FLUX.2-small-decoder and gemma-4-12B-it Models Now on Amazon SageMaker JumpStart

aiawsgadata-scientistaws-sagemaker
feature announcement

Amazon SageMaker JumpStart now includes Black Forest Labs' FLUX.2-small-decoder and Google's gemma-4-12B-it foundation models. This expansion offers AWS customers specialized capabilities for efficient image generation decoding and unified multimodal understanding. These additions enable the deployment of high-performance, scalable AI solutions, addressing diverse use cases with enhanced speed, lower memory consumption, and advanced agentic workflows. Customers can readily deploy these models via the SageMaker console or Python SDK.

  • →FLUX.2-small-decoder for Faster Image Generation
  • →gemma-4-12B-it for Multimodal Understanding and Agentic Workflows
Features (2) ›
  • FLUX.2-small-decoder for Faster Image Generation

    The FLUX.2-small-decoder is now available, optimized for faster image decoding and lower VRAM usage within FLUX.2 image generation pipelines. It offers approximately 1.4x faster decoding speed and 1.4x lower VRAM consumption as a distilled VAE decoder, making it suitable for production-grade image generation workloads at scale.

  • gemma-4-12B-it for Multimodal Understanding and Agentic Workflows

    Google's gemma-4-12B-it model has been added, excelling in unified multimodal understanding across text, image, and audio inputs. It natively supports function calling and agentic workflows, featuring an encoder-free architecture that delivers performance comparable to larger models with a smaller memory footprint, enabling powerful enterprise AI experiences.

Read the original announcement →

https://aws.amazon.com/about-aws/whats-new/2026/01/flux.2-small-decoder-gemma-4-12B-it-on-sagemaker-jumpstart/

Related releases