Overwhelming Demand Forces Kimi to Temporarily Halt New Subscriptions
AI assistant Kimi is facing a high-demand challenge. According to an official announcement, user demand for its Kimi K3 model has surged far beyond expectations in the last 48 hours. This unexpected traffic spike has pushed the platform's GPU computing resources close to their current capacity limits.
To safeguard the service quality for all existing paying subscribers, the Kimi team made a decisive move: immediately suspending new subscription sign-ups. This is a temporary measure, and all current subscribers will continue to receive full, uninterrupted service without any impact.
Computing Power Expansion Underway, Subscriptions to Reopen in Phases
Simply limiting access isn't a long-term solution. Kimi's announcement states that its technical team is working at full speed to scale up its computing infrastructure to address the root cause.
Once additional resources are securely integrated, the platform plans to reopen new subscription slots in controlled phases. This gradual approach aims to ensure that every new wave of users receives a quality experience, preventing another scenario of resource strain. Prospective users are advised to watch for further official updates.
Major Overhaul: Membership Splits into Two Tiers
Looking beyond the immediate capacity issue, Kimi shared a significant strategic shift for its service model. The platform's membership system will soon be divided into two distinct plans.
- Kimi Member: Designed for general users accessing the assistant via the web, mobile app, and Work scenarios. This plan covers everyday functions like conversation, document processing, and information retrieval.
- Kimi Code Member: A brand-new tier tailored specifically for developers and programming workflows. It will deeply integrate specialized features for code generation, debugging, and explanation, offering a more focused and powerful toolset for technical users.
Precision Operations for Long-Term Stability
The rationale behind splitting the membership is to enable more precise resource management and operations. By distinguishing between users with different needs and use cases, the platform can allocate computing power more accurately and optimize its service architecture.
For instance, the computational load patterns for generating complex code differ significantly from those of a simple Q&A session. A dedicated "Kimi Code" plan allows the platform to optimize resource pools and technical tuning specifically for programming tasks. This enhances efficiency for professional users while preventing their high-intensity tasks from affecting the smooth experience of general users.
This move signals Kimi's evolution from a general-purpose AI assistant towards a more specialized and segmented service model. As competition in the AI application space intensifies, product tiering to meet diverse user demands and ensure the long-term stability of core services has become a crucial strategy.