Qwen 1.5 0.5B Chat: benchmark results
Alibaba's Qwen1.5-0.5B-Chat, the aligned chat version of Qwen1.5-0.5B, post-trained with supervised fine-tuning and direct preference optimization, supporting a 32K context window. Provider: Alibaba. Released 2024-02-04. Access: Open.
Aggregate Rating 1192 ± 16, rank #1591 of 1656 rated models, from 161 benchmark results.
Strongest benchmark results
| Benchmark | Score | Metric | Percentile |
|---|---|---|---|
| RewardBench Chat Hard | 62.94 | Accuracy (%) | 54.5 |
| MERA - LCS | 11.8 | Accuracy (%) | 41.9 |
| EuroEval Norwegian Knowledge - Idioms NO | 1.21 | MCC (x100) | 38.4 |
| Open LLM Leaderboard - MuSR | 38.37 | Score | 31.4 |
| EuroEval Spanish NLU - MLQA ES | 44.87 | Character F1 (%) | 28.9 |
| Open LLM Leaderboard v1 - GSM8K | 7.66 | Accuracy (%) (5-shot) | 27.2 |
| EuroEval French NLU - FQuAD | 44.28 | Character F1 (%) | 27.1 |
| EuroEval Icelandic Knowledge | 2.15 | Knowledge Average Score (%) | 26.3 |
| EuroEval Portuguese NLU - MultiWikiQA PT | 41.92 | Character F1 (%) | 26.3 |
| EuroEval Italian NLU - SQuAD IT | 40.64 | Character F1 (%) | 25.1 |
| Open LLM Leaderboard - GPQA | 26.93 | Score | 23.8 |
| MERA - RWSD | 50.38 | Accuracy (%) | 22.1 |
Interactive version: theaggregate.ai/model?slug=qwen-1-5-0-5b-chat · How It Works · Data refreshed daily, snapshot 2026-10-20.