At midnight on July 19, Kimi announced that due to the rapid growth in user request volume, the current computing power cluster is approaching its capacity limit. The team has decided to suspend new consumer users from subscribing and has initiated a computing power expansion plan.

Kimi stated that in the past 48 hours, the user request volume has significantly exceeded previous estimates, and the current computing power resources are nearing the maximum capacity of the cluster. To ensure the service experience for existing subscribed users, Kimi will prioritize the allocation of existing computing power resources to ensure that the rights of subscribed users are not affected. Meanwhile, the team is accelerating the computing power expansion. Once the new computing power is gradually in place, more subscription slots will be gradually opened until normal subscription services are restored.

In terms of subscription system adjustments, Kimi plans to separate the benefits for subsequent new users, managing the "Kimi main benefit" and the "Kimi Code benefit" separately. The Kimi main benefit covers product services such as Kimi Web, Kimi APP, and Kimi Work, while Kimi Code is aimed at code-related usage scenarios, to achieve more precise matching of computing power resources and improve overall service stability.

This pause in new subscriptions reflects the computing power supply pressure faced by model service providers against the backdrop of rapid growth in AI applications. As the scale of large model applications continues to expand, the balance between computing infrastructure, resource scheduling capabilities, and user growth is becoming an important challenge for AI companies' commercial operations.

008Dh0eagy1if9ox2j9haj32s03pcqv8.jpg