DeepSeek will route its top model's users to a cheaper one.
The company said this week that V4.1 Flash, its budget model, beats V4 Pro on speed, cost and quality. From 10 September every V4 Pro request goes to the cheaper model, billed at the cheaper price. DeepSeek published no benchmark scores, and its public model list still shows no V4.1.
Why it mattersCheap Chinese models keep pushing down the price of work you may already pay for elsewhere. A sudden swap also breaks apps tuned to the old model, so your tools may behave differently.
Read in the app — free, in 9 languages
Related stories
Inception's model writes its whole answer at once, not word by word.2026-09-09A tiny open Chinese model matched one four times its size.
2026-09-09iFlytek trained a 293-billion-parameter model without Nvidia chips.
2026-09-07Outside testers scored OpenAI's new Astra below Anthropic's Fable 5.1.
2026-09-05A Chinese lab built Saudi Arabia's flagship Arabic AI model.
2026-09-04