China’s popular AI chatbot DeepSeek experienced its most extended outage since going viral in early 2025, going offline from late March 29 until the morning of March 30, 2026. The company’s official status page indicated the service was unavailable for 7 hours and 13 minutes, before being fully restored at 10:33 a.m. local time.
DeepSeek, known for its R1 and V3 models, has become a widely used tool for research, content creation, and coding assistance. During the outage, users were unable to access the web or app interfaces, frequently encountering error messages such as “server busy” or network prompts to retry. The disruption triggered a wave of online reactions, with the phrase “DeepSeek崩了” (“DeepSeek has crashed”) trending across Chinese social media.
The company provided updates on the status page but did not reveal the cause of the downtime. Analysts suggest possible explanations include software bugs, server failures, or system overloads, though no official confirmation was given. Despite the disruption, the outage is reported as resolved without lasting technical damage.
Many users noted interruptions to work and study routines, with some losing unsaved content during the offline period. Key features, including AI-driven reasoning and coding tools, remained inaccessible throughout the outage, highlighting the heavy reliance on the platform in daily tasks.
Prior to this incident, DeepSeek’s web interface had rarely been offline for more than two hours at a stretch, while its API service had experienced longer interruptions during high-demand periods in January 2025. The recent downtime has prompted renewed discussions about the robustness and resilience of large-scale AI systems amid growing global adoption.
Also Read: 12 tonnes of KitKat stolen in Europe transit