The API is also overwhelmed: DeepSeek experienced its sixth service disruption this month, with the impact now extending to the API for the first time.

According to 1M AI News monitoring, the DeepSeek status page indicates that on March 31 at 17:02 (CST), both the web interface and API service experienced performance issues simultaneously, with full recovery by 18:05, lasting approximately 1 hour. Previous incidents earlier this month, totaling five, were all labeled as "Web/App" or "Web" outages. This is the first time the API service has been included in the affected scope, indicating that not only front-end users but also third-party applications and developers accessing DeepSeek via the API were impacted.

This marks DeepSeek's sixth publicly documented service disruption in March. The status page history reveals that on March 5, the web interface was unavailable for about 45 minutes, on March 10, the web/app was unavailable for approximately 76 minutes, on March 18, a performance degradation lasted around 11 minutes, and on March 29-30, two consecutive incidents resulted in a total downtime of nearly 13 hours.