Model comparison
Gemma 2B vs Qwen-14B
Qwen-14B is the stronger model overall, scoring 30.8 to 29.1 on the Noometry Index.
Last verified . 17 shared benchmarks.
Google
Alibaba (Qwen)Summary
- They share 17 benchmarks with published results for both. Gemma 2B scores higher in 0 categories and Qwen-14B in 7 categories; 6 gaps are clear of the uncertainty.
- The widest gap is in multilingual, where Qwen-14B leads 27.7 to 23.1.
Side by side
| Gemma 2B | Qwen-14B | |
|---|---|---|
| Provider | Alibaba (Qwen) | |
| Noometry Index | 29.1 | 30.8 |
| Released | 2024-02-21 | 2023-09-24 |
| Weights | Open | Open |
| Context window | — | — |
| Max output | — | — |
| Input $ / M tokens | — | — |
| Output $ / M tokens | — | — |
| Results tracked | 23 | 18 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding Qwen-14B leads
Gemma 2B: 28.9 (#300), Qwen-14B: 30.6 (#283)
| Benchmark | Gemma 2B | Qwen-14B |
|---|---|---|
| LMArena Coding | 1010 | 1072 |
| HumanEval+ | 20.7% | — |
| MBPP+ | 34.1% | — |
Reasoning Too close to call
Gemma 2B: 17.9 (#275), Qwen-14B: 18.7 (#255)
| Benchmark | Gemma 2B | Qwen-14B |
|---|---|---|
| LMArena Hard Prompts | 989 | 1027 |
| BIG-Bench Hard | 35.2% | 55% |
| Epoch Capabilities Index | 93.57 | 112.8 |
| PIQA | 77.3% | 79.9% |
| HellaSwag | 71.4% | — |
| LAMBADA | — | 71.1% |
| WinoGrande | 65.4% | — |
Math Qwen-14B leads
Gemma 2B: 27.6 (#237), Qwen-14B: 28.7 (#224)
| Benchmark | Gemma 2B | Qwen-14B |
|---|---|---|
| LMArena Math | 1009 | 1068 |
| GSM8K | 17.7% | 61.3% |
Knowledge Not comparable
Gemma 2B: —, Qwen-14B: —
| Benchmark | Gemma 2B | Qwen-14B |
|---|---|---|
| ARC (AI2) Challenge | 42.1% | 84.4% |
| BoolQ | 69.4% | 86.2% |
| MMLU | 42.3% | 66.3% |
| TriviaQA | 53.2% | — |
Multilingual Qwen-14B leads
Gemma 2B: 23.1 (#289), Qwen-14B: 27.7 (#270)
| Benchmark | Gemma 2B | Qwen-14B |
|---|---|---|
| LMArena Non-English | 958 | 1042 |
| LMArena Chinese | 986 | 1077 |
| LMArena Russian | 937 | — |
Instruction Following Qwen-14B leads
Gemma 2B: 48.8 (#297), Qwen-14B: 52.6 (#284)
| Benchmark | Gemma 2B | Qwen-14B |
|---|---|---|
| LMArena Instruction Following | 971 | 1031 |
Long Context Qwen-14B leads
Gemma 2B: 29.9 (#286), Qwen-14B: 31.3 (#276)
| Benchmark | Gemma 2B | Qwen-14B |
|---|---|---|
| LMArena Longer Query | 981 | 1029 |
Writing & Preference Qwen-14B leads
Gemma 2B: 24.1 (#303), Qwen-14B: 27.8 (#294)
| Benchmark | Gemma 2B | Qwen-14B |
|---|---|---|
| LMArena Text | 1002 | 1051 |
| LMArena Creative Writing | 986 | 1028 |
| LMArena Multi-Turn | 945 | 1022 |
Frequently asked questions
Is Gemma 2B better than Qwen-14B?
Qwen-14B is the stronger model overall, scoring 30.8 to 29.1 on the Noometry Index.
Is Gemma 2B or Qwen-14B better for coding?
Qwen-14B scores higher on coding benchmarks: 30.6 versus 28.9 in the Noometry coding category.
How many benchmarks do Gemma 2B and Qwen-14B share?
17 benchmarks have published results for both models. Gemma 2B has 23 scored results on Noometry and Qwen-14B has 18.