aiqbench takeAI Products

Google DeepMind has released DiffusionGemma, an experimental 26B MoE model that leverages text diffusion to achieve up to 4x faster generation speeds on GPUs compared to standard autoregressive decoding. While the model prioritizes speed and supports parallel generation of non-linear text structures, it trades off some output quality, making it best suited for interactive local workflows rather than high-fidelity production tasks. The model runs efficiently within 18GB of VRAM and supports a 256K context window across 140+ languages, offering a distinct alternative for latency-sensitive applications.

Sources coveringMarkTechPost, arXiv

What this isA model's reading of the article and related coverage, written for aiqbench. Not the publisher's words.

Google AI Releases DiffusionGemma, a 26B MoE Open Model Using Text Diffusion for Up to 4x Faster Generation

MarkTechPostJune 10, 20268 min read

DiffusionGemma is Google DeepMind's experimental 26B open model using text diffusion for up to 4x faster generation on GPUs. The post Google AI Releases DiffusionGemma, a 26B MoE Open Model Using Text Diffusion for Up to 4x Faster Generation appeared first on MarkTechPost.

Google AI team including the Google DeepMind researchers have just released DiffusionGemma, an experimental open model for text generation. It uses text diffusion instead of standard autoregressive decoding. The model ships under a permissive Apache 2.0 license. Google positions it for devs and researchers exploring speed-critical, interactive local workflows. Examples include in-line editing, rapid iteration, and generating non-linear text structures. Most language models in use today are autoregressive. They generate one token at a time, left to right. Each new token depends on the token before it. DiffusionGemma works differently. It generates entire blocks of text simultaneously, in parallel. On dedicated GPUs, this delivers up to 4x faster generation. DiffusionGemma is a 26B Mixture of Experts (MoE) model. It activates only 3.8B parameters during inference. It is built on the Gemma 4 backbone, specifically the 26B-A4B architecture. Google integrated a diffusion…

Excerpt shown under the publisher's RSS terms · 141 words · full text and figures are on the publisher's site

Read at MarkTechPost ↗Ask about this article