Summary
The AI model quota did not reset after the scheduled weekly refresh period. In addition, the quota issue appears to affect all available models, even though I primarily used only Claude models.
The following message is displayed in the Models settings page:
“View your available model quota and AI credits. Model quota refreshes periodically based on your plan. Enable AI Credit Overages to continue using models when your quota is exhausted.”
However, the quota should have already refreshed today, and I have not used the new quota yet.
Expected Behavior
-
Model quotas should automatically refresh after the 7-day reset period.
-
All models should become available again after the reset.
-
Usage from one model family (e.g., Claude) should not incorrectly affect unrelated models.
Actual Behavior
-
The model quota does not refresh even after the reset period has passed.
-
All models appear to be affected by the quota limitation.
-
The system still behaves as if the previous quota cycle is active.
-
Models that were never used are also impacted.
Affected Models
-
Claude Sonnet 4.6 (Thinking)
-
Claude Opus 4.6 (Thinking)
-
Gemini 3.1 Pro
-
Gemini 3.1 Flash
-
GPT-OSS 120B
Additional Information
-
I primarily use Claude models only.
-
Today should have been the scheduled weekly quota reset date.
-
AI Credit Overages is enabled.
-
Available AI Credits are still shown as available (751 credits remaining).
-
Despite this, model quotas are not refreshing properly.
Steps to Reproduce
-
Use AI models until nearing the weekly quota limit.
-
Wait until the scheduled 7-day quota reset period.
-
Open Settings → Models.
-
Observe that the quota has not refreshed and all models remain affected.
Possible Cause
Potential causes may include:
-
Failed quota refresh synchronization.
-
Cached quota state not being cleared properly.
-
Shared/global quota calculation incorrectly affecting all models.
-
Backend quota refresh job not executing successfully.
