2.11
Queue priority per API key
Keys on Scale and Dedicated plans can mark runs as interactive. Interactive runs jump the batch queue.
Batch jobs and live users shared one queue. At peak hours a nightly import could slow down a chat window.
What changed
Keys on Scale and Dedicated plans can now set a priority. Interactive runs skip ahead of batch runs and wait on average 40 ms less at peak. Batch runs still finish; they just yield.