Apps & Tools
PlayableDiffusionGemma 26B-A4B-it-NVFP4
A quantized version of Google's DiffusionGemma 26B A4B model, optimized for NVIDIA GPUs.
Built with
DiffusionGemma 26B-A4BNEWModel credit: Strong — The page is the official Hugging Face repository for the model, which explicitly links to the base model card and details the architecture.
The model card explicitly states: 'Base Model: Gemma 4 26B A4B' and 'DiffusionGemma 26B A4B IT is an open-weights multimodal generative model developed by Google DeepMind'.Build evidence
Strong — This is a primary source repository hosted on Hugging Face with detailed documentation and model weights.
Shipped
2h ago · model from Jun 10, 2026This project provides an NVFP4 quantized version of the DiffusionGemma 26B A4B model. It leverages NVIDIA's Model Optimizer to enable high-speed multimodal text generation, supporting parallel 256-token block generation and massive context lengths on NVIDIA Hopper and Blackwell hardware.
#quantization#multimodal#nvidia#model-optimization
Timeline
Teaser
Demo
Playable
Product
Loading…