BANGLAPLAN: A Benchmark for Constraint-Based Planning and Decision-Making in Bangla

Researchers introduced BANGLAPLAN, a benchmark of 1,000 constraint-based planning and decision-making problems in Bangla spanning ten task families and four language registers, according to an arXiv paper (arXiv:2610.19704v1). Evaluating 11 open-weight and proprietary model configurations, the team found GPT-6.1 Sol reached 98.7% accuracy and Qwen3.5-27B with thinking reached 80.3%, while direct-answer configurations scored at most 30.9%, with seven of eleven configurations falling below the always-"infeasible" baseline. The benchmark, available at https://github.com/mrayhanulmasud/banglaplan, reports feasible-case accuracy, infeasibility detection, and format errors separately, and the authors say harder problems mainly reduce plan-construction accuracy while infeasibility detection stays more robust.

arXiv:2610.19704v1 Announce Type: new Abstract: Large language models LLMs increasingly help users arrange appointments, plan journeys, and allocate limited resources. These tasks require models to satisfy constraints on time, cost, capacity, and order, or determine that no feasible solution exists. Bangla is the world's seventh most widely spoken language, and existing benchmarks cover many language and reasoning tasks. However, benchmarks for evaluating planning and decision-making in Bangla under explicit constraints remain lacking. We introduce BANGLAPLAN, a benchmark of 1,000 problems across ten task families grounded in everyday life and institutions in Bangladesh. It includes feasible and infeasible problems in four language registers: everyday, formal, colloquial, and Bangla--English mixed. Answers cover eight structured types, including schedules, routes, allocations, and ordered steps, in a common JSON format. We construct the benchmark through language-model-assisted drafting and independent human validation. We evaluate 11 open-weight and proprietary model configurations using exact match, separately reporting feasible-case accuracy, infeasibility detection, and format errors. We find a large performance gap: GPT-6.1 Sol achieves 98.7% accuracy and Qwen3.5-27B with thinking reaches 80.3%, while direct-answer configurations score at most 30.9%. Seven of eleven configurations perform below the always- infeasible'' baseline. Our findings also suggest that harder problems primarily reduce plan-construction accuracy, while infeasibility detection remains more robust. BANGLAPLAN provides a reproducible basis for evaluating planning and decision-making in Bangla. BANGLAPLAN is available at: https://github.com/mrayhanulmasud/banglaplan.