Model Limitation: Need Claude Opus 4.6–Level Conversational Quality

The current output quality also feels limited by the underlying model. The responses lack the natural conversational ability and flow needed for real interviews. A model with higher conversational capability (at the level of Claude Opus 4.6 or equivalent) would significantly improve the quality and make responses feel more human and usable in live scenarios.