The use of large language models (LLMs) for complex mathematical reasoning is an emergent area of research, with fast progress in methods, models, and benchmark datasets. However, most mathematical reasoning evaluations exhibit a significant linguistic bias, with the vast majority of benchmark datasets being exclusively in English or (at best) translated from English. We address this limitation by introducing {\sc Math-PT}, a novel dataset comprising 1,729 mathematical problems written in European and Brazilian Portuguese. {\sc Math-PT} is curated from a variety of high-quality native sources, including mathematical Olympiads, competitions, and exams from Portugal and Brazil. We present a comprehensive benchmark of current state-of-the-art LLMs on {\sc Math-PT}, revealing that frontier reasoning models achieve strong performance in multiple choice questions compared to open weight models, but that their performance decreases for questions with figures or open-ended questions. To facilitate future research, we release the benchmark dataset and model outputs.
@article{arxiv.2604.25926,
title = {MATH-PT: A Math Reasoning Benchmark for European and Brazilian Portuguese},
author = {Tiago Teixeira and Ana Carolina Erthal and Juan Belieni and Beatriz Canaverde and Diego Mesquita and Miguel Faria and Eliezer de Souza da Silva and André F. T. Martins},
journal= {arXiv preprint arXiv:2604.25926},
year = {2026}
}
Comments
Accepted at 17th International Conference on Computational Processing of Portuguese (PROPOR 2026). Open access to dataset repo https://huggingface.co/datasets/tiagoteixeira03/MATH-PT and model outputs https://github.com/deep-spin/math-benchmark