After experiencing a nearly 12-hour outage on March 29th to 30th, DeepSeek went down again around 5 p.m. on March 31st, and it wasn’t until after 6 p.m. in the evening that the service resumed normal operation. “DeepSeek Goes Down” once again became a hot topic for discussion that day.
On the afternoon of March 31st, many users took to social media to report that the DeepSeek system once again encountered abnormal access issues, frequently displaying messages such as “Please check your network and try again” or “Server is busy” during interactions.
Netizens complained one after another, “Is this the artificial intelligence that was once worshiped as a god?”,”What’s wrong now, my dear! I’m working on my assignment here, and you’re crashing, causing chaos!”,”Again, it’s trying to take my life, now I can’t get into my class.”,”Can it be fixed by tomorrow? I haven’t even written my material today.”,”How many times does it need to crash in a day to be considered over?”,”It just crashes every now and then.”
According to a report by Interface News, the service status page of the DeepSeek system on that day showed that at 5:02 p.m., the DeepSeek webpage and functions experienced abnormalities. It wasn’t until 6:05 p.m. in the evening that the related services were restored.
Prior to this, on the evening of March 29th to the morning of March 30th, DeepSeek had just experienced a nearly 12-hour crash.
At around 9:35 p.m. on March 29th, the DeepSeek webpage and app services showed abnormalities, causing disruptions for many users who were rushing to write their papers, debug their code, or complete their projects. Some users lost unsaved content due to the service interruption. Finally, on the morning of March 30th, the system was repaired, setting a new record for the longest single crash in DeepSeek history.
Historical data shows that DeepSeek has experienced multiple service disruptions in the past.
On March 10th this year, DeepSeek services had a major abnormal access incident. In March, May, and July of last year, DeepSeek also made headlines several times due to system failures.
It’s not just DeepSeek, as reported by Interface News previously, the current AI industry is facing a severe shortage of computing power, with many AI platforms frequently experiencing crashes due to insufficient computational resources. Kimi also crashed previously due to a shortage of computing power.