Source item

didn’t think this would happen for a while, but now I consistently downgrade model for chat queries I know are not that hard, should be mostly retrieval/ai search api bc I want faster speed cc: @KeenableAI