Model comparison
Qwen1.5-14B vs StarCoder 2 3B
Qwen1.5-14B has enough public results to be ranked (#250); StarCoder 2 3B does not yet, so treat this comparison as directional.
Last verified . 1 shared benchmarks.
Alibaba (Qwen)
NVIDIASummary
- They share 1 benchmark with published results for both. Qwen1.5-14B scores higher in 0 categories and StarCoder 2 3B in 1 category, but none of those gaps is larger than the uncertainty.
Side by side
| Qwen1.5-14B | StarCoder 2 3B | |
|---|---|---|
| Provider | Alibaba (Qwen) | NVIDIA |
| Noometry Index | 32.1 | 35.8 |
| Released | 2024-02-04 | 2024-02-22 |
| Weights | Open | Open |
| Context window | — | — |
| Max output | — | — |
| Input $ / M tokens | — | — |
| Output $ / M tokens | — | — |
| Results tracked | 17 | 7 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding Too close to call
Qwen1.5-14B: 32.5 (#259), StarCoder 2 3B: 32.9
| Benchmark | Qwen1.5-14B | StarCoder 2 3B |
|---|---|---|
| LMArena Coding | 1138 | — |
| BigCodeBench Complete | — | 21.4% |
| HumanEval+ | — | 27.4% |
Reasoning Not comparable
Qwen1.5-14B: 20.5 (#217), StarCoder 2 3B: —
| Benchmark | Qwen1.5-14B | StarCoder 2 3B |
|---|---|---|
| LMArena Hard Prompts | 1113 | — |
| Epoch Capabilities Index | — | 88.01 |
| WinoGrande | — | 57.1% |
Math Not comparable
Qwen1.5-14B: 29.9 (#214), StarCoder 2 3B: —
| Benchmark | Qwen1.5-14B | StarCoder 2 3B |
|---|---|---|
| LMArena Math | 1125 | — |
| GSM8K | — | 21.6% |
Knowledge Not comparable
Qwen1.5-14B: 29.3 (#230), StarCoder 2 3B: —
| Benchmark | Qwen1.5-14B | StarCoder 2 3B |
|---|---|---|
| MMLU | 68.6% | 36.6% |
| LMArena Expert | 1094 | — |
| ARC (AI2) Challenge | — | 34.2% |
Multilingual Not comparable
Qwen1.5-14B: 30.8 (#257), StarCoder 2 3B: —
| Benchmark | Qwen1.5-14B | StarCoder 2 3B |
|---|---|---|
| LMArena Non-English | 1095 | — |
| LMArena Chinese | 1147 | — |
| LMArena French | 1115 | — |
| LMArena German | 1043 | — |
| LMArena Japanese | 1019 | — |
| LMArena Russian | 1046 | — |
| LMArena Spanish | 1085 | — |
Instruction Following Not comparable
Qwen1.5-14B: 57.1 (#266), StarCoder 2 3B: —
| Benchmark | Qwen1.5-14B | StarCoder 2 3B |
|---|---|---|
| LMArena Instruction Following | 1102 | — |
Long Context Not comparable
Qwen1.5-14B: 33.7 (#252), StarCoder 2 3B: —
| Benchmark | Qwen1.5-14B | StarCoder 2 3B |
|---|---|---|
| LMArena Longer Query | 1113 | — |
Writing & Preference Not comparable
Qwen1.5-14B: 33.7 (#270), StarCoder 2 3B: —
| Benchmark | Qwen1.5-14B | StarCoder 2 3B |
|---|---|---|
| LMArena Text | 1128 | — |
| LMArena Creative Writing | 1091 | — |
| LMArena Multi-Turn | 1110 | — |
Frequently asked questions
Is Qwen1.5-14B better than StarCoder 2 3B?
Qwen1.5-14B has enough public results to be ranked (#250); StarCoder 2 3B does not yet, so treat this comparison as directional.
Is Qwen1.5-14B or StarCoder 2 3B better for coding?
They score almost the same on coding (32.5 vs 32.9); test both on your own repository before choosing.
How many benchmarks do Qwen1.5-14B and StarCoder 2 3B share?
1 benchmark has published results for both models. Qwen1.5-14B has 17 scored results on Noometry and StarCoder 2 3B has 7.