The outage happened in the early hours of March 30th and wasn't resolved until around 10:33 AM (local time in China). And that fucked up the things I had to do during that time window :( The company didn't explain the cause, but issues like this usually stem from server failures or bugs after updates. This is notable because, until now, the chatbot had never been offline for so long since it went viral in 2025 with the R1 and V3 models. Interestingly, at the height of its popularity in 2025, the API had already experienced long outages, but the public interface had never been down for more than ~2 hours. https://www.reuters.com/technology/chinas-deepseek-ai-chatbot-suffers-longest-outage-since-viral-rise-early-2025-2026-03-30/


Did you forget that there's something way better? [claude](https://claude.ai/new)