Large Language Models (LLMs) have revolutionized business communication with automated email writing, improved efficiency, and personalization. This paper presents a comparative study of four widely used AI-powered LLMs in business email writing considering the following parameters: formality, readability, grammatical correctness, and word limit. This research compares models such as OpenAI-GPT-4o-mini, Google-Gemini-2.0-Flash, Claude-3.7-Sonnet, and Meta-Llama3-70b in producing professional business emails that fit various scenarios. This work quantitatively compares AI-produced emails by adopting automated scoring. Based on our findings, Google-Gemini-2.0-Flash leads its peers in producing professional, polished, and grammatically correct business communication. This research aims to help professionals choose the most appropriate AI LLM to write effective and contextually relevant emails.