Bangladesh Legal QA (No Law): leaderboard
Metric: Accuracy (%; no legal context supplied, parametric knowledge only; over 150 supplied-law control MCQs, strict consistency (correct in all four option orders); reference instruction-tuned model). Source: arxiv.org. Saturation forecast: Around December 2026. 6 models tracked.
Top models
| # | Model | Score |
|---|---|---|
| 1 | Qwen 3.5 4B | 76.7 |
| 2 | Gemma 4 E2B | 70 |
| 3 | Qwen 3.5 2B | 57.3 |
| 4 | Llama 3.2 3B | 50.7 |
| 5 | Qwen 3.5 0.8B | 38.7 |
| 6 | Llama 3.2 1B | 28.7 |
Interactive version: theaggregate.ai/benchmark?slug=bangladesh-legal-qa-no-law · How It Works · Data refreshed daily, snapshot 2026-09-26.