Qwen2.5 Coder 1.5B — Benchmarks
Benchmark scores for Qwen2.5 Coder 1.5B aggregated from public leaderboards, with how it ranks among open models. See hardware requirements for what you need to run it.
Overall rank: #71 of 123 open modelscomposite 43.8/100 across 5 benchmarks in 3 categories · methodology
Knowledge
| Benchmark | Score | Open rank | All models |
|---|---|---|---|
| HellaSwag | 61.8% | #35 / 42 | #62 / 76 |
| ARC Challenge | 45.2% | #35 / 51 | #53 / 77 |
| MMLU | 53.6% | #55 / 77 | #105 / 136 |
Math
| Benchmark | Score | Open rank | All models |
|---|---|---|---|
| GSM8K | 65.8% | #21 / 60 | #30 / 93 |
Reasoning
| Benchmark | Score | Open rank | All models |
|---|---|---|---|
| WinoGrande | 60.7% | #37 / 46 | #71 / 80 |
Scores aggregated from public benchmark sources (each linked from the benchmark pages). llmrun does not run these benchmarks.