Qwen 1.5 72B Chat: benchmark results
Alibaba's 72B Qwen1.5 chat model (February 2024) with 32K context and strong multilingual coverage. Provider: Alibaba. Released 2024-02-05. Access: Open.
Unified ELO 1496 ± 1, rank #707 of 1392 rated models, from 50 benchmark results.
Strongest benchmark results
| Benchmark | Score | Metric | Percentile |
|---|---|---|---|
| Open Chinese LLM - C-Eval Semantic | 91.14 | Accuracy (%) | 98.2 |
| Open Chinese LLM - HellaSwag | 72.16 | Accuracy (%) | 96.4 |
| Open Chinese LLM Leaderboard | 68.83 | Average Score (%) | 93.2 |
| Open Chinese LLM - ARC Challenge | 62.12 | Accuracy (%) | 91.8 |
| Open Chinese LLM - CMMLU | 70.92 | Accuracy (%) | 89.3 |
| Open Chinese LLM - WinoGrande | 68.9 | Accuracy (%) | 88.3 |
| BiGGen-Bench | 3.83 | Average Score (1-5) | 85.3 |
| Open Chinese LLM - GSM8K | 59.67 | Accuracy (%) | 83.7 |
| Open PL LLM - Multiple Choice | 58.92 | Average Multiple-Choice Score (%) | 81.3 |
| MAGI-Hard | 63.47 | Accuracy (%, 2024-05 snapshot) | 79.3 |
| AlpacaEval 2.0 | 36.57 | LC Win Rate (%) | 76.6 |
| Open PL LLM Leaderboard | 58.67 | Average Score (%) | 74.5 |
Interactive version: theaggregate.ai/model?slug=qwen-1-5-72b-chat · How It Works · Data refreshed daily, snapshot 2026-09-05.