Swallow - Pre-trained English - OpenBookQA Leaderboard · The Aggregate

Swallow - Pre-trained English - OpenBookQA: leaderboard

Metric: Accuracy (%). Source: swallow-llm.github.io. 50 models tracked.

Top models

#	Model	Score
1	Llama 3.1 70B	45
2	Llama 4 Scout Base	43.2
3	Qwen 2.5 72B	41.6
4	Qwen3-14B-Base	41.6
5	gemma-3-27B-pt	41.4
6	Qwen3-30B-A3B-Base	41.4
7	Gemma 2 27B	41.2
8	Qwen 2.5 14B	41.2
9	Qwen 2.5 32B	40.6
10	gemma-3-12B-pt	39.8
11	Qwen 2.5 7B	39.2
12	sarashina2-70B	38.8
13	Gemma 2 9B	38.2
14	Qwen3-8B-Base	38.2
15	Qwen3-4B-Base	38.2

Interactive version: theaggregate.ai/benchmark?slug=swallow-pre-trained-english-openbookqa · How It Works · Data refreshed daily, snapshot 2026-09-19.