Google’s DiffusionGemma is an Apache 2.0 experimental open model using text diffusion instead of standard autoregressive decoding. The 26B MoE model activates 3.8B parameters during inference and is designed for low-latency local workflows. Google claims up to 4x faster generation on dedicated GPUs, while noting that output quality is below standard Gemma 4 and production-quality use cases should still prefer Gemma 4.
Google released DiffusionGemma, a 26B MoE experimental open model using text diffusion instead of token-by-token autoregressive decoding. It can generate blocks of text in parallel, reaching up to 4x faster output on dedicated GPUs. The model targets local, speed-sensitive workflows, but Google says its output quality is below standard Gemma 4 and recommends Gemma 4 for quality-critical production use.