Gemma

Amazon Web Services today announced availability of Google DeepMind's Gemma 4 family on Amazon Bedrock. The open-weight models include three instruction-tuned variants spanning 2.3B to 30.7B parameters, with 256K context windows, multimodal input support, and built-in reasoning mode.

Google DeepMind released DiffusionGemma, a 26B parameter mixture-of-experts model that generates text using discrete diffusion instead of autoregression. The model processes blocks of 256 tokens in parallel, achieving generation speeds exceeding 1100 tokens per second on H100 GPUs in low-batch settings.

NVIDIA has released Nemotron 3.5 Content Safety, a 4B-parameter model built on Google Gemma 3 4B IT that provides multimodal safety classification across approximately 140 languages. The model includes a 128K context window, custom enterprise policy enforcement, auditable reasoning traces, and is releasing its training dataset.

Google launched AI Edge Gallery for macOS, allowing Mac users to run Google's Gemma models locally. The platform ships with five Gemma models, including the newly released Gemma 4 12B—a 12-billion-parameter multimodal model that handles text, vision, and audio while running on consumer laptops with 16GB of RAM.

Three new models appeared on Hugging Face: Ideogram AI's FP8-quantized version of its Ideogram-4 image generation model and Google's Gemma-4-12B text models in both base and instruction-tuned variants. The releases mark continued expansion of model availability through Hugging Face's platform.

Google DeepMind released the Gemma 4 E4B assistant model using Multi-Token Prediction (MTP) architecture that accelerates inference by up to 2x through speculative decoding. The 4.5B effective parameter model supports 128K context windows and handles text, image, and audio input with pricing not yet disclosed.

NVIDIA engineer Asier Arranz demonstrated Gemma 4 running as a vision-language agent (VLA) on a Jetson Orin Nano Super with 8GB RAM. The model autonomously decides when to access a webcam based on user queries, with no hardcoded triggers—performing speech-to-text, vision analysis, and text-to-speech entirely locally.

NVIDIA released Nemotron-Personas-Korea, a dataset containing 7 million demographically accurate synthetic personas grounded in official Korean statistics from KOSIS, Supreme Court of Korea, and the National Health Insurance Service. The dataset includes 26 fields per persona covering demographics, geography, and occupation across all 17 Korean provinces, with zero personally identifiable information under CC BY 4.0 license.

Google released Gemma 4, an open-weights model family that runs directly on edge devices with multi-step planning and autonomous workflow capabilities. The Apache 2.0 licensed model bypasses traditional cloud security controls by executing entirely on local hardware, creating a governance blind spot for enterprise security teams.