DeepSeek experienced two service outages in one night, spanning nearly 13 hours, which even trended on Weibo.

According to 1M AI News's monitoring, the DeepSeek status page indicates that on the night of March 29 to the morning of March 30, both the web and app experienced two separate incidents of service disruption, which have now been resolved.

The first "service anomaly" began at 9:35 p.m. on March 29, with the team identifying the issue at 10:28 p.m. and restoring service by 11:23 p.m., lasting approximately 1 hour and 48 minutes. Only 57 minutes later, at 12:20 a.m. on March 30, the second "performance anomaly" occurred. After an initial fix at 1:24 a.m. that failed after 52 minutes, a second fix at 9:13 a.m. eventually brought full recovery by 10:33 a.m., resulting in about a 10-hour incident. During the incident, "DeepSeek went down" trended on Weibo, with users reporting seeing "server busy" messages and being unable to load content in new conversations.