# Evaluating Multimodal LLMs for Inpatient Diagnosis: Real-World Performance, Safety, and Cost Across Ten Frontier Models

arXiv: https://arxiv.org/abs/2604.16980 · HTML: https://mlsanity.com/paper/2604.16980/

## Key facts

- No citing papers in the map yet (data through 14 Oct 2026).
- Builds on 7 of its 24 references that are in the map.
- First posted on arXiv 18 Apr 2026.

## Builds on (most cited first, 7 of 7)

- [Judging LLM-as-a-Judge with MT-Bench and Chatbot Arena](https://mlsanity.com/paper/2306.05685/) (2023)
- [Can Large Language Models Be an Alternative to Human Evaluations?](https://mlsanity.com/paper/2305.01937/) (2023)
- [Adapted Large Language Models Can Outperform Medical Experts in Clinical Text Summarization](https://mlsanity.com/paper/2309.07430/) (2023)
- [Capabilities of Gemini Models in Medicine](https://mlsanity.com/paper/2404.18416/) (2024)
- [Sequential Diagnosis with Language Models](https://mlsanity.com/paper/2506.22405/) (2025)
- [A prospective clinical feasibility study of a conversational diagnostic AI in an ambulatory primary care clinic](https://mlsanity.com/paper/2603.08448/) (2026)
- [Can LLMs Accurately Score Medical Diagnoses and Clinical Reasoning?](https://mlsanity.com/paper/2604.14892/) (2026)

## Abstract (arXiv)

Background: Large language models (LLMs) are increasingly proposed for diagnostic support, but few evaluations use real-world multimodal inpatient data, particularly in low and middle-income country (LMIC) public hospitals. Methods: We conducted VALID, a retrospective evaluation of 539 multimodal inpatient cases from a tertiary public hospital in South Africa. Inputs included radiology imaging (CT, MRI, CXR) and reports, laboratory results, clinical notes, and vital signs. Expert panels adjudicated 300 cases (balanced and discordant subsets) to establish ground truth diagnoses, differentials, and reasoning. Ten multimodal LLMs generated zero-shot outputs. A calibrated three-model LLM Jury scored all outputs and routine ward diagnoses across diagnostic accuracy, differential quality, reasoning, and patient safety (>10,000 evaluations). Primary outcomes were composite scores (S3, S4) and win rates. Results: (i) LLM performance was tightly clustered (<15% variation) despite large cost differences; low-cost models performed comparably to top models. (ii) All LLMs significantly outperformed routine ward diagnoses on average diagnostic and safety scores. (iii) Top performance was achieved by GPT-5.1, followed by Gemini models. (vi) Adding radiology reports improved performance by 6%. (v) Diagnostic and reasoning scores were highly correlated (ρ = 0.85). (vi) Output rates varied (65-100%) due to input constraints. Results were robust across subsets and evaluation design. Conclusions: Across a real-world LMIC dataset, multimodal LLMs showed similar diagnostic performance despite large cost differences and outperformed routine care on average safety metrics. Affordability, robustness, and deployment constraints may outweigh marginal performance differences in LMIC settings.