Back to papers
March 19, 2026cs.IRIntermediate

Comparative Analysis of Large Language Models in Generating Telugu Responses for Maternal Health Queries

AI-Generated Summary

This study compares how three popular AI chatbots (ChatGPT-4o, Gemini, and Perplexity) answer pregnancy-related health questions in Telugu, a language spoken in India. Using both automatic scoring and expert reviews from gynecologists, the researchers found that Gemini performed best overall, while Perplexity did well when questions were asked in Telugu rather than English. The findings highlight that both which AI tool you use and what language you ask in significantly impact the quality of health information provided in non-English languages.

Difficulty
Intermediate
Categories

cs.IR

AI Tags
Large Language ModelsTelugu LanguageHealthcareMaternal HealthMultilingual NLPLow-Resource LanguagesModel EvaluationComparative Analysis