Use GPT-4o-mini for applying

Feature Requests
1

Would be great to use the cheap and fast GPT-4o-mini applying in composer cause of how fast it is and how big its token output can be