# confidence-based decision fallback

AI Gateway can now escalate a decision request to a fallback model when the primary model's answer trips a confidence condition the user sets. Conditions can be combined, so one escalation can depend on more than one signal; the plain model-name fallbacks already in use keep catching outright errors and are unaffected. Confidence conditions cover Choice and Score questions, Boolean questions escalate on a probability range, and a triggered fallback runs a second decision, so both stages are billed.

AI Gateway can now escalate a decision request to a fallback model when the primary model's answer trips a confidence condition the user sets. Conditions can be combined, so one escalation can depend on more than one signal; the plain model-name fallbacks already in use keep catching outright errors and are unaffected. Confidence conditions cover Choice and Score questions, Boolean questions escalate on a probability range, and a triggered fallback runs a second decision, so both stages are billed.