What's New¶

Recent updates across InferFoundry products.

This page is updated weekly. Check back every Friday for the latest changes.

Week of August 18, 2026¶

InferFoundry API

• Budget alerts and spend tracking: The API now notifies you when your team's spend crosses budget thresholds, and the spend dashboard includes your current period's spend, available budget, and maximum budget at a glance.
• Daily activity endpoints: New endpoints let you view daily token usage broken down by model for your keys, teams, and users, including cache token usage. Date ranges default to the last 30 days if not specified.
• Model access groups: Administrators can now define model access groups to control which models specific teams or keys can use.
• Model aliases and EOL dates: The public models list now shows model aliases, end-of-life dates, and whether prompt caching is enabled (with accurate cache pricing when it is).
• Region selection when creating teams: You can now choose a region directly from the create-team dialog.
• Key last-used timestamp by region: A new endpoint lets you look up when an API key was last used in each region.
• Disposable email blocking: Sign-ups using disposable email addresses are now blocked to improve account quality.
• Email normalization: Email addresses are now normalized consistently during login and registration, preventing duplicate accounts from case or formatting differences.
• Security – key isolation: Fixed a bug where individually-owned keys could be visible to other members of the same team.
• Security – key access control: Added additional safeguards to prevent accessing another team's keys by ID.
• Vendor model name display: Fixed model display names to preserve the vendor's original casing.
• Billing resilience: The billing system now recovers gracefully from sync failures during top-ups and billing cycle transitions.
• Subscription cancellation: When a subscription is cancelled, any remaining budget is now allocated fairly using a first-in, first-out approach.