[DeepSeek V4.1 Flash Model Officially Released] DeepSeek has officially released the DeepSeek V4.1 Flash model. This is the smallest model in the company's new series of model architectures, featuring native multimodal visual understanding capabilities. According to the introduction, the new generation model significantly reduces the size of the KV Cache. Compared to the previous generation model, it reduces HBM requirements to one-fourth and SSD requirements to one-eighth. In Agent usage scenarios, cache hit costs often account for a high proportion, and the compression of the KV Cache greatly lowers the usage costs for Agent-related tasks.