Availability metrics are reported at an aggregate level across all tiers, models, and error types. Individual customer availability may vary depending on their subscription tier (PAYG, Scale-Tier, or Reserved Capacity), the specific model and API features in use, as well as differences in the user's network environment.