DeepSeek Releases DSpark for Accelerated LLM Per-User Generation

DeepSeek has released DSpark, a speculative decoding framework designed to accelerate large language model inference in production environments. DSpark utilizes a parallel draft backbone with a tiny sequential head to improve token acceptance rates and includes a confidence head and load-aware scheduler to dynamically adjust token verification based on GPU workload. In production, DSpark speeds up per-user generation on DeepSeek-V4 models by 60-85% compared to their previous single-token setup, without compromising output quality.