Quota Accuracy & Faster Loads
Version: 1.19.18, Date: November 10, 2025
Quota accuracy is back on track in this release. Users who saw library or prompt counts showing incorrectly (like a Libraries card at 0 despite having dozens) will find those numbers now reflect reality. O1 model credit costs were also being calculated at the wrong rate and have been corrected, and quota and usage pages load noticeably faster. The rest of this post is grouped by app area.
Settings
- Fixed: The quota display now shows correct resource counts. The Libraries stat card could show 0 even if you had dozens of libraries. The system now accurately tracks and displays your total libraries, prompts, and versions using a more reliable counting method.
- Fixed: Credit deductions for OpenAI’s O1 models were being calculated at the wrong rate. Your credit balance now accurately reflects actual usage when using these models.
- Improved: The quota and usage pages load noticeably faster. The underlying data now fetches in parallel, so checking your usage feels snappier.
Across the App
- Improved: Behind the scenes, persistent resources (libraries, prompts, versions) are now tracked separately from monthly-resetting resources (credits). This provides a more reliable foundation for accurate usage reporting going forward.

