IndicKLAR (Bengali): leaderboard

Metric: Accuracy (%; relaxed prefix match of the answer to 2,619 factual queries in Bengali script, native-speaker verified; direct 3-shot prompting with English examples, greedy decoding). Source: arxiv.org. Saturation forecast: Estimated already saturated. 9 models tracked.

Top models

#ModelScore
1Gemma 3 12B (IT)74.1
2Llama 3.1 8B Instruct58.4
3Qwen 2.5 14B Instruct55.4
4Gemma 3 4B (IT)52.5
5Qwen 2.5 7B Instruct44
6Llama 3.2 3B Instruct40.3
7Gemma 3 1B (IT)28.7
8Qwen 2.5 1.5B Instruct23.4
9Llama 3.2 1B Instruct19.6

Interactive version: theaggregate.ai/benchmark?slug=indicklar-bengali · How It Works · Data refreshed daily, snapshot 2026-09-26.