- Introduced the new [Flex](https://ai.google.dev/gemini-api/docs/flex-inference) and [Priority](https://ai.google.dev/gemini-api/docs/priority-inference) inference tiers, offering more options for optimizing cost or latency.