DeepSeek has officially launched its new V4.1 Flash model as of September 10, 2026. The company has introduced a specific pricing structure for this model, which includes a cache hit input price of ¥0.02, a cache miss input price of ¥1, and an output price of ¥4.
The company also announced a schedule change regarding its V4 Pro service. The retirement of the V4 Pro service has been postponed to September 14, 2026, at 12:00 Beijing time. After this deadline, all requests directed to V4 Pro will be automatically routed to the V4.1 Flash model.
Following the transition, users will be billed according to the V4.1 Flash pricing scheme for all routed requests. Internal and external testing results indicate that the new V4.1 Flash model outperforms the V4 Pro version across key metrics, including performance, processing speed, and total execution time.
