Summary
Cloudflare Workers AI introduced a rejectIfBusy option for synchronous inference that makes requests fail immediately when capacity is unavailable, rather than waiting in a queue. It can be passed as a third argument to the AI binding or in the REST request body.
What changed
Workers AI added a rejectIfBusy option for synchronous inference; when set, requests fail immediately instead of waiting in a capacity queue.
Why it matters
Latency-sensitive applications often prefer a fast failure they can route around over an unbounded queue wait; giving developers explicit backpressure control makes Workers AI more predictable under load and easier to build fallbacks against.
Evidence excerpt
Added rejectIfBusy option for synchronous Workers AI inference, allowing requests to fail immediately when capacity unavailable rather than queue. "Use it when your application should not wait in a capacity queue."