Imagine dictating complex equations and having them instantly transformed into perfectly formatted LaTeX code. That's the promise of Greek2MathTex, a groundbreaking new AI system that translates spoken Greek mathematics into ready-to-use LaTeX. This innovative tool opens doors for greater accessibility in STEM, particularly for visually impaired individuals who often face barriers with LaTeX's intricate syntax. The system combines cutting-edge speech recognition with the power of large language models (LLMs) like GPT-3.5. It works by first transcribing the spoken Greek mathematical expression into text. Then, a clever retrieval mechanism finds similar equations and their corresponding LaTeX from a custom-built dataset called Gr2Tex. These examples, along with carefully crafted instructions (prompts), are fed to the LLM, which then generates the LaTeX code for the dictated equation. The researchers behind Greek2MathTex meticulously fine-tuned their system, experimenting with different LLMs, prompting strategies, and similarity measures to achieve optimal performance. They even developed a custom evaluation metric to ensure accuracy. The results are impressive, with the system accurately translating a high percentage of spoken equations into LaTeX. This technology isn't just a research project; it's available as a user-friendly web application. Users can simply record their equation, listen back, and then click a button to generate the LaTeX code, which appears in a convenient pop-up window. While initially focused on Greek, the underlying technology holds immense potential for other languages and could revolutionize how we interact with mathematical notation. Future research aims to explore even more advanced AI models and prompting techniques, like Chain-of-Thought prompting, to further improve accuracy and expand the system's capabilities. Greek2MathTex is more than just a tool; it's a step toward a more inclusive and accessible future for STEM.