ZCode 3.8.1 adds off-peak reset accessOfficial release note describing a manual reset mechanism, not a global automatic reset.Official ZCode changelog
Source ↗THE RESET OBSERVATORY
When will GLM reset?
Explore GLM Coding Plan reset cards, ZCode usage rules, and public reset records. Check card access and expiry; manual resets are separate from global resets.
Latest collected snapshot · 2026-09-10 10:31 UTC
CodexHistory availableClaudeHistory availableGeminiUsage rulesGLMReset cardsKimiReset recordsAntigravitySeparate quota
Loading collected public records…
GLM reset watch
GLM Coding Plan. Check your BigModel or Z.ai account for the benefits that apply to it.
Check for an available reset card
ZCode supports cards that manually reset five-hour or weekly quotas. Off-peak grants depend on timing, usage and daily limits; they are not scheduled or guaranteed for every account.
Read the official guide ↗Historical coverage is insufficient for a percentage forecast.
Evidence reviewed 2026-09-10Rule check pending or failed Sep 10, 2026, 10:31 AM UTC
Reviewed public records
2Event days usable for forecasting
0Official rule sources
0 / 1Evidence review date
2026-09-10THE RECORD
Public evidence and reset records
Reported quota reset at the GLM-5.3 launchMedia relay of the launch announcement. The original announcement was not retrieved; excluded from forecast samples.Jiemian report via Eastmoney
Source ↗FOLLOW THE SIGNAL
Public sources, clearly identified.
Codex and Claude use third-party reset catalogs. Other platforms show reviewed public records and official rules. Periodic rule checks do not discover new announcements. Regular renewals, manual reset cards and policy changes are kept separate from automatic-reset forecasts.
GLM reset-card guide
StaleOfficial rules · Periodic checks, not announcement discovery
Last success Sep 10, 2026, 10:31 AM UTC
Open source ↗HOW WE ESTIMATE
An estimate,
not a promise.
Platforms can change their policies at any time. These estimates do not mean your account will receive a reset.
01Start with history, then consider new signals
Over the past 90 days, automatic reset announcements on the same UTC date count as one event day. We need at least 30 days of observation and 5 event days to show a probability. The third-party historical record may be incomplete.
The Gamma–Poisson model uses α = 1 + event days and β = 7 + observed days. The baseline probability of at least one announcement in the next h days is 1 − (β / (β + h))^α.
02How public signals affect the estimate
Reset previews, model releases, usage incidents, and community signals have manually chosen weights of 2.0, 0.35, 0.25, and 0.10. Each is scaled by source weight and decays with age. Only the past 72 hours count, and only the strongest signal in each group is used.
These weights are experimental assumptions, not trained or calibrated values. Banked resets and ordinary client updates do not increase the automatic-reset estimate. Without valid new signals, the adjusted estimate equals the baseline. The 7-day result is a longer-range extrapolation.
03How we evaluate forecasts
No probability forecasts have been issued for this platform. Usage rules, manual reset cards and relayed reports do not create forecast samples or prospective scores.
Historical replay and prospective scoring are kept separate. We need timestamped forecasts and comparisons against the baseline using Brier score and Log Loss before claiming better accuracy.
04Open-source references and event definitions
tibo-reset-watch informed public-post monitoring; tibo-reset-lab provides an auditable forecasting experiment. This site is independently implemented.
Automatic resets, banked resets that must be redeemed, and regular account usage renewals are different events. Read OpenAI’s explanation ↗