Asale

Plans & capacity

What each subscription tier is worth on the market, and how Asale estimates how much of it is still free.

Asale never asks a provider "how much quota is left" — none of them will say. It estimates instead, from the plan you are on and what this device has already served.

Rolling window capacity

Most subscriptions meter over a rolling 5-hour window. These are the estimates Asale declares:

ProviderPlan5h windowFull-day equivalent
ClaudeMax 20×4,400,00021.1M
ClaudeMax 5×2,200,00010.6M
ClaudeTeam / Enterprise1,500,0007.2M
ClaudePro1,100,0005.3M
ChatGPT / CodexPro2,000,0009.6M
ChatGPT / CodexPlus1,000,0004.8M
GeminiUltra3,000,00014.4M
GeminiPro / Advanced1,500,0007.2M
Kimi, Grokany500,0002.4M

Figures are tokens. The full-day column is the window scaled by ×4.8 — it is a reference figure for setting a percentage cap, not a limit anyone enforces.

An unrecognised plan falls back to the lowest paid tier for that provider. Under-declaring is the safe direction: over-declaring sends the market work your provider then refuses, which costs reputation.

Kimi, Grok, and API keys

Moonshot and xAI express their limits as requests over a window, not tokens, and their platform APIs are metered against a balance rather than capped. There is nothing to estimate, so the window is set conservatively and your daily cap is the control that matters.

A custom endpoint has no window at all — it is a metered key against somebody's balance. The daily cap is the only bound.

Which plan should sell?

If you haveThen
Max 20× and use a fraction of itThe best supply on the network. Cap at 50–70% and forget it.
Pro, used dailySell overnight, buy during the day. Cap the sell limit so your own working hours are never the ones that get squeezed.
No subscriptionBuy only. No monthly fee, no minimum — you pay per token.
Several accountsEach is a separate switch and cap. Sell the idle ones; leave the working one's Buy switch off so it stays direct.

Reading your own usage

Usage and Limits in the client show, per account: served this window, served today, remaining, and whether you are on pace to run out before the window resets.

Some providers expose a live limits endpoint and the bars are exact. Where they do not, the bars are estimated from what this device has metered — which is why the client says estimate rather than pretending otherwise.