# Do Multilingual VLMs Reason Equally? A Cross-Lingual Visual Reasoning Audit for Indian Languages

arXiv: https://arxiv.org/abs/2603.26742 · HTML: https://mlsanity.com/paper/2603.26742/

## Key facts

- No citing papers in the map yet (data through 14 Oct 2026).
- Placed in Reasoning benchmarks and evaluation, one of 17 directions in LLM reasoning.
- Builds on 23 of its 27 references that are in the map.
- First posted on arXiv 23 Mar 2026.

## Placed in

- LLM reasoning / [Reasoning benchmarks and evaluation](https://mlsanity.com/lr/reasoning-benchmarks-and-evaluation/)
- Multimodal LLMs / vision-language models / [VLM benchmarks and visual reasoning](https://mlsanity.com/mllm/vlm-benchmarks-and-visual-reasoning/)

## Builds on (most cited first, 23 of 23)

- [Qwen2.5-VL Technical Report](https://mlsanity.com/paper/2502.13923/) (2025)
- [Qwen3-VL Technical Report](https://mlsanity.com/paper/2511.21631/) (2025)
- [InternVL: Scaling up Vision Foundation Models and Aligning for Generic Visual-Linguistic Tasks](https://mlsanity.com/paper/2312.14238/) (2023)
- [MMMU: A Massive Multi-discipline Multimodal Understanding and Reasoning Benchmark for Expert AGI](https://mlsanity.com/paper/2311.16502/) (2023)
- [Learn to Explain: Multimodal Reasoning via Thought Chains for Science Question Answering](https://mlsanity.com/paper/2209.09513/) (2022)
- [MathVista: Evaluating Mathematical Reasoning of Foundation Models in Visual Contexts](https://mlsanity.com/paper/2310.02255/) (2023)
- [Language Models are Multilingual Chain-of-Thought Reasoners](https://mlsanity.com/paper/2210.03057/) (2022)
- [Measuring Multimodal Mathematical Reasoning with MATH-Vision Dataset](https://mlsanity.com/paper/2402.14804/) (2024)
- [MEGA: Multilingual Evaluation of Generative AI](https://mlsanity.com/paper/2303.12528/) (2023)
- [Visually Grounded Reasoning across Languages and Cultures](https://mlsanity.com/paper/2109.13238/) (2021)
- [IndicTrans2: Towards High-Quality and Accessible Machine Translation Models for all 22 Scheduled Indian Languages](https://mlsanity.com/paper/2305.16307/) (2023)
- [CVQA: Culturally-diverse Multilingual Visual Question Answering Benchmark](https://mlsanity.com/paper/2406.05967/) (2024)
- [Towards Leaving No Indic Language Behind: Building Monolingual Corpora, Benchmark and Models for Indic Languages](https://mlsanity.com/paper/2212.05409/) (2022)
- [xGQA: Cross-Lingual Visual Question Answering](https://mlsanity.com/paper/2109.06082/) (2021)
- [Pangea: A Fully Open Multilingual Multimodal LLM for 39 Languages](https://mlsanity.com/paper/2410.16153/) (2024)
- [All Languages Matter: Evaluating LMMs on Culturally Diverse 100 Languages](https://mlsanity.com/paper/2411.16508/) (2024)
- [Aya Vision: Advancing the Frontier of Multilingual Multimodality](https://mlsanity.com/paper/2505.08751/) (2025)
- [DRISHTIKON: A Multimodal Multilingual Benchmark for Testing Language Models' Understanding on Indian Culture](https://mlsanity.com/paper/2509.19274/) (2025)
- [Chitrarth: Bridging Vision and Language for a Billion People](https://mlsanity.com/paper/2502.15392/) (2025)
- [IndicVisionBench: Benchmarking Cultural and Multilingual Understanding in VLMs](https://mlsanity.com/paper/2511.04727/) (2025)
- [The Reasoning Lingua Franca: A Double-Edged Sword for Multilingual AI](https://mlsanity.com/paper/2510.20647/) (2025)
- [VIRAASAT: Traversing Novel Paths for Indian Cultural Reasoning](https://mlsanity.com/paper/2602.18429/) (2026)
- [VLURes: Benchmarking VLM Visual and Linguistic Understanding in Low-Resource Languages](https://mlsanity.com/paper/2510.12845/) (2025)

## Abstract (arXiv)

Vision-language models score well on mathematical, scientific, and spatial reasoning benchmarks, yet these evaluations are overwhelmingly English. I present the first cross-lingual visual reasoning audit for Indian languages. 980 questions from MathVista, ScienceQA, and MMMU are translated into Hindi, Tamil, Telugu, Bengali, Kannada, and Marathi using IndicTrans2, with Gemini 2.0 Flash cross-verification on 50 samples per language (inter-translator agreement 0.79-0.84). Eight VLMs, from 7B open-source models to GPT-4o, are evaluated across all seven languages, yielding 68,600 inference records that include text-only and chain-of-thought ablations. I find accuracy drops of 9.8-25 percentage points when switching from English to an Indian language, with Dravidian languages suffering up to 13.2 pp more than Indo-Aryan. Chain-of-thought prompting degrades Bengali (-14.4 pp) and Kannada (-11.4 pp) rather than helping, exposing English-centric reasoning chains. Aya-Vision-8B, built for 23 languages, still drops 28.5 pp on Dravidian scripts; multilingual pretraining alone does not transfer visual reasoning. I release the translated benchmark and all model outputs.