IndicConBench: A Multilingual Constitutional Reasoning Benchmark for India

IndicConBench is a multilingual constitutional reasoning benchmark for India designed to evaluate Large Language Models. It assesses factual accuracy, reasoning capacity, and legal intelligence on the Constitution of India in English and Hindi. The dataset was developed by Vikhram S and published under Vikhram Labs.

Use Cases

• Benchmarking LLM factual accuracy on the Indian Constitution based on the described evaluation focus.
• Evaluating multilingual reasoning capacity in English and Hindi as specified in the description.
• Assessing structural retrieval capability of legal text from a defined corpus.
• Testing legal intelligence of models on constitutional questions.

Strengths

• Designed as a publication-grade, research-quality benchmark.
• Specifically targets evaluation of LLMs on the Constitution of India.
• Supports multilingual analysis in both English and Hindi.

Limitations

• Column-level documentation is absent; field semantics must be inferred after download.
• Row count is unknown, which may limit suitability assessment.
• The dataset's freshness should be verified as the last update timestamp is 2026-05-30.

Provenance

Source

Vikhram Labs (author: vikhram-labs).

Freshness

Last updated 2026-05-30 09:53:24.

Geography

India