Gemini Diffusion vs Step-3.5-Flash

Gemini Diffusion

Google DeepMind

Gemini Diffusion is an experimental text and code generation model from Google DeepMind, announced at Google I/O in May 2025 as the first diffusion-based language model to achieve quality comparable to autoregressive models on standard benchmarks. Unlike transformer-based models that predict tokens sequentially left-to-right, it generates entire blocks of text by iteratively refining noise — the paradigm used in image and video generation models — enabling faster sampling speeds and stronger mid-generation error correction for code and mathematical editing tasks. At announcement it was available only as an experimental demo via waitlist, with no public API, marking it as a research milestone rather than a production deployment.

StepFun

Step-3.5-Flash, released by StepFun on February 2, 2026, is a Mixture-of-Experts large language model with 197 billion total parameters and approximately 11 billion active parameters per inference. It features a 256K token context window using a 3:1 sliding-window-to-full-attention ratio, processing 100–350 tokens per second. Step-3.5-Flash targets agentic tasks, coding workflows, and open-source deployments requiring frontier reasoning capabilities with efficient inference, under an Apache 2.0 license.