Routing Strategy

Core Routing Principle

To ensure enterprise-grade reliability and compliance control, TokenHub routing follows a fixed order: User Compliance Requirements → Provider Availability → Cache Consistency. In practice, requests satisfy compliance boundaries first, then route among available providers for stable and consistent delivery.

Scope of User Preference Constraints

You can configure region, provider, and model constraints in user preferences. Once enabled, subsequent requests route only within the configured boundaries, supporting unified governance across compliance, procurement, and performance goals.

Availability Protection and Auto Failover

The platform continuously observes global provider availability. When service disruption or quality degradation is detected, traffic fails over to the next available provider. Per request, the backend attempts up to 3 providers to maximize success rate and continuity.

In wide-area incidents (such as major network infrastructure failures), short-term unavailability may occur. In such cases, retry later.