DeepSeek formally launches V4.1 Flash, routes V4 Pro requests to Flash

DeepSeek formally released the V4.1 Flash model on September 10, introducing updated Flash-series pricing for API users. According to company notifications, internal and external testing indicates that V4.1 Flash outperforms V4 Pro regarding performance, cost, speed, and total completion time. Consequently, starting September 14 at noon Beijing time, the company will begin routing requests originally intended for V4 Pro to the new V4.1 Flash model to optimize efficiency and delivery for its users.