Sber Open-Sources GFusion, Russia's First Diffusion Language Model
Sber
Sber released GFusion, an experimental diffusion-based language model built on GigaChat3-10B-A1.8B. Unlike autoregressive models, GFusion sketches a structural outline then fills tokens in parallel passes (~32 tokens per pass). Internal benchmarks show 45–70% faster generation than GigaChat 3 at a cost of 2–4 percentage points of quality. Weights published on Hugging Face alongside custom TileLang attention kernels and SGLang integration.
Why it matters
First Russian open-source diffusion LLM, placing Sber alongside Google (Diffusion Gemma) and Inception Labs in the emerging non-autoregressive generation category.
Importance: 3/5
First Russian open-source diffusion language model; confirmed by official Sber Habr post and three independent Russian media outlets