A recent study has demonstrated GPT-4V’s superiority over Gemini Pro by utilizing visual question answering techniques for educational settings. The study’s highlights include:

• Quantitative and qualitative analyses were conducted using a dataset from student-drawn scientific models.
• GPT-4V significantly outperforms Gemini Pro in scoring accuracy and Quadratic Weighted Kappa.
• The adaptation of the NERIF method was unable to give Gemini Pro any advantage over GPT-4V.

This evidence puts GPT-4V ahead in the race of AI advancements, especially for educational tools involving multimodal data.

The importance of this study cannot be overstated. GPT-4V’s ability to outperform Gemini Pro in handling intricate educational tasks reveals substantial potential for AI applications in education. This opens doors for further research in AI-driven educational tools, potentially transforming how learning and assessments are approached.