IFEval#20 of 28
prompt_level_strict_acc
View run →76.5%±1.8%
Evaluated across 13 benchmarks. Ranks in the top 3 on 0 of 13. Strongest showing on IFEval (76.5% prompt_level_strict_acc, #20 of 28). Weakest on MBPP(Instruct) (0.0% pass@1, #7 of 24).
| Benchmark | Metric | Score | Rank | Actions |
|---|---|---|---|---|
| IFEval | prompt_level_strict_acc | 76.5%±1.8% | #20 of 28 | View run → |
| BBH | exact_match | 72.7%±0.5% | #4 of 28 | View run → |
| MMLU | acc | 67.2%±0.4% | #16 of 28 | View run → |
| EQ-Bench | eqbench | 65.0±2.7 | #23 of 28 | View run → |
| GSM8K | exact_match | 62.4%±1.3% | #20 of 28 | View run → |
| MBPP | pass@1 | 58.0%±2.2% | #11 of 28 | View run → |
| MMLU-Pro | exact_match | 55.7%±0.4% | #19 of 28 | View run → |
| MGSM | exact_match | 55.5%±0.8% | #21 of 28 | View run → |
| GPQA Main | acc | 33.0%±2.2% | #9 of 28 | View run → |
| GPQA Diamond | acc | 32.3%±3.3% | #9 of 28 | View run → |
| LongBench | aggregate | 31.9%±0.4% | #5 of 17 | View run → |
| GPQA Extended | acc | 31.7%±2.0% | #8 of 28 | View run → |
| MBPP(Instruct) | pass@1 | 0.0%±0.0% | #7 of 24 | View run → |
Citation
FrozeBench. "Qwen/Qwen3-4B-AWQ." https://frozebench.com/models/qwen-qwen3-4b-awq. Retrieved 2026-06-04.
BibTeX
@misc{frozebench_Qwen_Qwen3_4B_AWQ,
title = {Qwen/Qwen3-4B-AWQ},
howpublished = {\url{https://frozebench.com/models/qwen-qwen3-4b-awq}},
year = {2026},
note = {FrozeBench. Retrieved 2026-06-04.}
}URL
https://frozebench.com/models/qwen-qwen3-4b-awq