IndicIFEval: A Benchmark for Verifiable Instruction-Following Evaluation in 14 Indic Languages

TL;DR

IndicIFEval is a new benchmark designed to evaluate the performance of large language models in following instructions across 14 Indic languages, addressing a critical gap in multilingual NLP evaluation.

Contribution

The paper introduces IndicIFEval, a comprehensive benchmark with verified examples for 14 Indic languages, enabling systematic evaluation of LLMs' instruction-following capabilities in these languages.

Findings

Models perform well on formatting but struggle with lexical and cross-lingual tasks.

Progress in high-resource Indic languages is notable, but overall performance lags behind English.

IndicIFEval and scripts are publicly released to foster further research.

Abstract

Instruction-following benchmarks remain predominantly English-centric, leaving a critical evaluation gap for the hundreds of millions of Indic language speakers. We introduce IndicIFEval, a benchmark evaluating constrained generation of LLMs across 14 Indic languages using automatically verifiable, rule-based instructions. It comprises around 800 human-verified examples per language spread across two complementary subsets: IndicIFEval-Ground, translated prompts from IFEval (Zhou et al., 2023) carefully localized for Indic contexts, and IndicIFEval-Ground, synthetically generated instructions grounded in native Indic content. We conduct a comprehensive evaluation of major open-weight and proprietary models spanning both reasoning and non-reasoning models. While models maintain strong adherence to formatting constraints, they struggle significantly with lexical and cross-lingual tasks -- and despite progress in high-resource languages, instruction-following across the broader Indic family lags significantly behind English. We release IndicIFEval and its evaluation scripts to support progress on multilingual constrained generation (http://github.com/ai4bharat/IndicIFEval).