Qwen 2 1.5B: benchmark results
Alibaba's Qwen2-1.5B base language model, part of the Qwen2 series spanning 0.5 to 72 billion parameters plus a Mixture-of-Experts variant, using grouped-query attention. Provider: Alibaba. Released 2024-06-06. Access: Open.
Aggregate Rating 1324 ± 16, rank #1433 of 1656 rated models, from 133 benchmark results.
Strongest benchmark results
| Benchmark | Score | Metric | Percentile |
|---|---|---|---|
| Pinocchio Italian - Lingua Straniera | 72.31 | Accuracy (%) | 75 |
| Open LLM Leaderboard v1 - GSM8K | 55.8 | Accuracy (%) (5-shot) | 67.5 |
| Open LLM Leaderboard v1 - MMLU | 57.26 | Accuracy (%) (5-shot) | 40.8 |
| Open Chinese LLM - C-Eval Semantic | 65.58 | Accuracy (%) | 39.9 |
| MERA - RWSD | 54.23 | Accuracy (%) | 39.8 |
| Open LLM Leaderboard - MATH Level 5 | 7.02 | Score | 37.5 |
| Open Chinese LLM - CMMLU | 50.6 | Accuracy (%) | 37 |
| Open Japanese LLM - Xlsum JA Rouge1 | 23.25 | Score (%) | 36.3 |
| Open Japanese LLM - Xlsum JA RougeLsum | 19.65 | Score (%) | 35.1 |
| Open PL LLM - PolEmo2-IN (generative, 5-shot) | 69.81 | Accuracy (%) | 34.2 |
| MERA - ruHumanEval | 3.78 | pass@1 (%) | 34 |
| Open LLM Leaderboard v1 - TruthfulQA MC2 | 45.92 | MC2 (%) (0-shot) | 32.8 |
Interactive version: theaggregate.ai/model?slug=qwen-2-1-5b · How It Works · Data refreshed daily, snapshot 2026-10-20.