TECH Signal 479
DeepSeek V4.1 Flash launch set for September 10, Pro requests billed at Flash rates
↗
Illustration only Photo by Ben Vaughn on Unsplash
DeepSeek will release V4.1 Flash around September 10, 2026, with all Pro requests routed to it and billed at Flash pricing until V4.1 Pro arrives.
For engineers using DeepSeek's API, this means the Pro model they call will actually be served by the cheaper, faster V4.1 Flash, and they'll pay Flash rates. The pricing change takes effect at noon Beijing time on September 10, with off-peak rates of $0.003 per input cache hit, $0.15 per input cache miss, and $0.6 per output token, doubling during peak hours. This could significantly reduce costs for high-volume workloads, but also means behavior may differ from the previous Pro model.
Written by elseif from the cluster below · every claim links back to a sourceThe three things worth knowing
V4.1 Flash is scheduled for official release around September 10, 2026, and reportedly surpasses V4 Pro on performance, cost, speed, and task completion time.
Until V4.1 Pro launches, all requests to the Pro model will be routed to V4.1 Flash and billed at Flash's price.
Flash series pricing changes at noon Beijing time on September 10, with off-peak input cache hits at $0.003, cache misses at $0.15, and output at $0.6, doubling during peak hours.
THE CLUSTER