Multimodal medical question answering (MMQA) is a vital area bridging healthcare and Artificial Intelligence (AI). This survey methodically examines the MMQA research published in recent years. We collect academic literature through Google Scholar, applying bibliometric analysis to the publications and datasets used in these studies. Our analysis uncovers the increasing interest in MMQA over time, with diverse domains such as natural language processing, computer vision, and large language models contributing to the research. The AI methods used in multimodal question answering in the medical domain are a prominent focus, accompanied by applicability of MMQA to the medical field. MMQA in the medical field has its unique challenges due to the sensitive nature of medicine as a science dealing with human health. The survey reveals MMQA research to be in an exploratory stage, discussing different methods, datasets, and potential business models. Future research is expected to focus on application development by big tech companies, such as MedPalm. The survey aims to provide insights into the current state of multimodal medical question answering, highlighting the growing interest from academia and industry. The identified research gaps and trends will guide future investigations and encourage collaborative efforts to advance this transformative field.

Benchmark-driven evaluation helps distinguish between planning quality and interface reliability when large language models are utilized for embodied reasoning in simulation. Our submission to the Embodied Agent Interface Challenge (EAI) is evaluated across four stages of the pipeline. These being goal interpretation, subgoal decomposition, action sequencing, and transition modeling. The tasks run in the BEHAVIOR and VirtualHome simulators, which use constrained action vocabularies, fixed-object inventories and symbolic state representations within a standard evaluation protocol. Our system accesses the OpenAI API using GPT-4.1 for BEHAVIOR, GPT-4.1-mini for VirtualHome, and GPT-5-mini in later exploratory experiments across both environments. The schemas for each task determine how the outputs are structured, and outputs are regenerated when they do not follow the specification. On the final public leaderboard, our system ranked eighteenth overall with a score of 57.92, achieving 68.88 on BEHAVIOR and 46.96 on VirtualHome. In this paper, we describe our approach and discuss what these observations suggest about the strengths and limitations of current language models when used for embodied reasoning.