MOONSHOT · TEXT MODEL
kimi-k3 API access through TokenAAS
A planned Kimi language model entry for next-generation conversational and agent workloads.
- Agent applications
- Advanced chat
- Knowledge assistance
- Unified endpoint: /v1/chat/completions
- Billing: per million input and output tokens
Frequently asked questions
How do I call kimi-k3 through TokenAAS?
Create a TokenAAS API key and send the model name kimi-k3 to the /v1/chat/completions endpoint.
How is kimi-k3 usage billed?
Usage is billed per million input and output tokens. The live model catalog shows the currently configured TokenAAS rate.
Is kimi-k3 always available?
Availability is capacity-aware and may change. Check the live model catalog before production use and implement retry or fallback handling.