DeepSeek to adjust flash series pricing from Sept 10

This pricing adjustment signals a strategic push to shape demand patterns in AI inference, likely aiming to smooth load by incentivizing off-peak usage. For enterprises and developers reliant on DeepSeek's flash series, the cost structure shift could alter deployment economics, particularly for high-volume, cache-heavy workloads. The doubling of peak rates is the key lever to watch—whether it successfully redistributes traffic or merely raises the bill for time-sensitive applications remains an open question. Competitors may also recalibrate pricing in response, making this a potential inflection point for AI infrastructure cost dynamics.