A fallback chain is an ordered list of backup models. If the primary model returns a retryable error - a provider 5xx, a rate limit, or a timeout - the gateway automatically retries with the next model in the chain. Fallbacks are how a single provider outage stops becoming an application outage.
Why it matters
Provider outages and rate limits are routine. Without a fallback, a single upstream failure becomes your application's failure.
How it works
You list backup models in order. On a retryable error - 5xx, rate limit, or timeout - the gateway retries the next model automatically. Logs show which model actually served each request.