DeepSeek launches V4.1 Flash model, outperforms V4 Pro with new pricing

DeepSeek has emailed API users to announce the official launch of its V4.1 Flash model and new Flash pricing. After internal and external testing, V4.1 Flash now beats V4 Pro across performance, cost, speed, and total time.

The planned shutdown of V4 Pro has been pushed to 12:00 Beijing time on September 14, 2026. After that, V4 Pro will route to V4.1 Flash and be billed at the new rate. The new pricing takes effect at 12:00 on September 10: off-peak input cache hit is 0.02 yuan, cache miss is 1 yuan, and output is 4 yuan. Peak-hour prices are double the off-peak rates.

DeepSeek's open platform has also sent similar notices to users.