On October 15, DeepSeek experienced a major failure that impacted its web and mobile applications. The issue began at 08:11 PM local time, leaving users unable to send messages in the chat. For those relying on the platform, the interruption persisted for over three hours as the system struggled with service unavailability.
Reports suggest the cause of the failure may have been related to computational power issues. Neural networks demand significant resources, and the disruption highlighted the challenges of maintaining stability when such power demands fluctuate.
Last week some of South Korea's biggest banks were hit by a cyberattack. Thanks to a report from CrowdStrike tonight, we now know the entire hack may have been done by a single person. He used a combined stack of an open-source AI penetration tool named ARTEX, DeepSeek v4.1-Flash, GLM-5.3, Grok 4.6, and Claude Code.
The impact was widespread, affecting both the web chat and API services. Users attempting to engage with the platform found themselves unable to send messages, while developers relying on the API faced similar connectivity hurdles during the outage period.
DeepSeek identified the issue and rolled out a fix to address the problems with chat services and APIs, including those for DeepSeek V4 Pro and V4.
DeepSeek identified the issue and rolled out a fix to address the problems with chat services and APIs, including those for DeepSeek V4 Pro and V4.1 Flash. By 12:49 p.m., the company had moved the incident to monitoring status, signaling that the corrective measures were in place to restore functionality.
DeepSeek implements fix after partial outage affects chat and API services https://t.co/nM9cCIc2kC
The timeline confirms that the service was impacted for over three hours. The rapid identification of the issue and the subsequent deployment of a fix allowed the company to stabilize its infrastructure and resume normal operations for its users.
This incident serves as a reminder of the operational demands placed on large-scale neural networks. By addressing the fault lines in its chat and API architecture, DeepSeek demonstrated its capacity to manage service interruptions and restore access to its core tools.
DeepSeek identifies cause of outage; fix underway After 08:00 Beijing time on Oct 15 DeepSeek’s website and app showed "server temporarily unavailable." The company said a partial web/API outage affected the Chatservice dialogue service, DeepSeek V4.1 Flash API (APIService) and one other component; the cause has been identified and fixes are being implemented. (https://t.co/Rr1vBK6ruR)