AI chatbot Kimi tests credit-based spending limits for heavy users
The move reflects a broader shift among AI companies toward usage-based pricing as compute costs climb.
3-MIN READ
By Pacific Post Staff·Published
A smartphone displays the Kimi AI chatbot branding · Source: Reuters
Moonshot AI's chatbot Kimi has begun experimenting with a credit-based spending system that caps how much computing power heavy users can draw on, part of a broader shift across the AI industry toward usage-based pricing models.
The approach ties a user's spending more directly to the computing resources their queries consume, a departure from simple flat-rate subscriptions. As AI models have grown more capable, the computing cost of running them has climbed sharply, pushing developers to look for pricing structures that better reflect actual usage.
Industry watchers say the shift mirrors trends already underway at other AI companies, many of which have introduced tiered plans or usage quotas as the cost of running large language models continues to strain budgets. For users, the change can mean more transparency about what a given query costs, but also a new layer of complexity in managing usage.
The experiment comes as competition among Chinese AI developers intensifies, with several companies racing to expand their user bases while managing the steep infrastructure costs that come with offering increasingly powerful models. How companies balance accessibility with sustainable pricing is likely to remain a central question as the sector matures.
