On this page

Start with /usage

Kimi's help material points CLI users to `/usage` for current benefits and quota. Check it before a long task and after the task finishes. The change is more informative than a raw count of prompts.

If you use Kimi through several official clients, confirm whether the membership allowance is shared across those surfaces. The account-level view matters more than one local session log.

Model speed can change quota use

Kimi's model documentation notes that some faster options can consume quota at a higher multiplier. A speed increase may be worth it when time matters, but it should be a deliberate tradeoff rather than an invisible default.

Record the model, task, and quota change. You will quickly learn which work benefits from high speed and which work is fine on the regular option.

Keep your usage visible while you work. Get Super Notchy for $27 ↗

Keep reset windows in view

Super Notchy reads the local Kimi Code CLI session and can show the short rate window and weekly quota when available. That keeps the reset timing beside other assistants on the Mac.

Use the official Kimi view for account detail and the persistent ring for pacing. If one window becomes tight, switch models, narrow the task, or route work to another assistant with available capacity.

Sources and further reading