* Introduced the new [Flex](/docs/flex-inference) and [Priority](/docs/priority-inference) inference tiers, offering more options for optimizing cost or latency.
