DeepSeek crashes again: Website and model API both malfunction, just 19 days after a nearly 5-hour outage.

Beating AI News Flash, November 3, 16:58 — DeepSeek experienced another service outage. The official status page showed partial disruptions across web chat, V4 Pro API, and V4.1 Flash API. As of 17:06, the official team was still investigating and had not yet disclosed the specific cause.

This comes just 19 days after the last large-scale outage on October 15. At that time, the same three services were affected together, and the outage lasted nearly 5 hours. DeepSeek subsequently issued a public apology, announced that API users would receive bonus credits equal to their consumption amount over the previous 7 days, and committed to a comprehensive review and the elimination of single points of dependency in the system.