Large Language Models (such as Chat GPT) are currently experiencing great popularity in the financial industry. The strengths and weaknesses of these models are demonstrated on the basis of two realized prototypes. Valuable insights for further implementations can be derived from this.