DeepSeek experiences major service outage affecting web and API

The Chinese artificial intelligence platform DeepSeek experienced a significant service outage on October 15, affecting its web interface, mobile application, and various API services, including DeepSeek V4.1 Flash. The disruption began around 8:11 AM Beijing time and lasted for approximately five hours, impacting users and developers during peak morning working hours.

Users reported various issues, such as messages failing to receive replies, loading screens hanging, and error messages stating the “server is temporarily unavailable.” The outage had global reach, with reports of slow or non-responsive services from users in locations including New York, Poland, and Luxembourg.

DeepSeek confirmed it had identified the cause and implemented repairs, eventually announcing service restoration by early afternoon. To compensate affected API users, the company announced it would issue account credits equivalent to the consumption amounts recorded between October 8 and October 14.

Industry analysts noted that this is not an isolated incident, as the platform has faced multiple service interruptions throughout the year. Experts suggest these frequent outages may stem from a mismatch between computing power supply and demand, or insufficient operational maintenance to handle rapid expansion.