DeepSeek service was "partially interrupted" for half an hour, with a total of five failures in May, sounding the reliability alarm.

On the morning of May 28, the DeepSeek service failed, and web conversations and APIs were affected. According to the official status page, the fault severity level is "partial outage" - the problem was located at 10:21, resolved and restored at 10:50, and the interruption lasted for about half an hour. This is DeepSeek's fifth service issue in May 2026, two of which were "complete outages."

The frequency of five failures per month, in the context of the explosive growth of V4 series calls (OpenRouter weekly calls 6.75 trillion Tokens, V4 Flash single week 3.65 trillion in the last week of May, a month-on-month increase of 32%), exposed the deep contradiction that the expansion speed of DeepSeek engineering infrastructure cannot keep up with the growth of model demand. For production users relying on the DeepSeek API, five failures in a month has hit a reliability redline. Solving the computing power bottleneck is not technically difficult - just add servers and purchase more cloud computing power - but it means higher daily operating expenses, and DeepSeek's B-side strategy is to continue to lower prices to grab users. How to find a balance between "price butcher" and "service stability" is the most pressing challenge facing the DeepSeek engineering team.