Model comparison
Llama 2-70B vs Qwen-14B
Qwen-14B is the stronger model overall, scoring 31.4 to 24.4 on the Noometry Index.
Last verified . 18 shared benchmarks.
Meta
Alibaba (Qwen)Summary
- They share 18 benchmarks with published results for both. Llama 2-70B scores higher in 5 categories and Qwen-14B in 2 categories; 4 gaps are clear of the uncertainty.
- The widest gap is in math, where Qwen-14B leads 31.2 to 8.1.
Side by side
| Llama 2-70B | Qwen-14B | |
|---|---|---|
| Provider | Meta | Alibaba (Qwen) |
| Noometry Index | 24.4 | 31.4 |
| Released | 2023-07-18 | 2023-09-24 |
| Weights | Open | Open |
| Context window | — | — |
| Max output | — | — |
| Input $ / M tokens | — | — |
| Output $ / M tokens | — | — |
| Results tracked | 35 | 18 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding Too close to call
Llama 2-70B: 31.4 (#286), Qwen-14B: 31.2 (#288)
| Benchmark | Llama 2-70B | Qwen-14B |
|---|---|---|
| LMArena Coding | 1079 | 1071 |
Reasoning Qwen-14B leads
Llama 2-70B: 14.4 (#325), Qwen-14B: 19.6 (#257)
| Benchmark | Llama 2-70B | Qwen-14B |
|---|---|---|
| LMArena Hard Prompts | 1073 | 1027 |
| BIG-Bench Hard | 64.9% | 55% |
| Epoch Capabilities Index | 113.79 | 113.03 |
| LAMBADA | 78.9% | 71.1% |
| PIQA | 82.8% | 79.9% |
| DTBench | 41.6% | — |
| CommonsenseQA 2.0 | 50% | — |
| ForecastBench | 51.4 | — |
| HellaSwag | 85.3% | — |
| WinoGrande | 80.2% | — |
Math Qwen-14B leads
Llama 2-70B: 8.1 (#326), Qwen-14B: 31.2 (#227)
| Benchmark | Llama 2-70B | Qwen-14B |
|---|---|---|
| LMArena Math | 1091 | 1068 |
| GSM8K | 69.6% | 61.3% |
| OTIS Mock AIME 2024-2025 | 0% | — |
| MATH Level 5 | 3.3% | — |
Knowledge Not comparable
Llama 2-70B: 7.4 (#310), Qwen-14B: —
| Benchmark | Llama 2-70B | Qwen-14B |
|---|---|---|
| ARC (AI2) Challenge | 78.3% | 84.4% |
| BoolQ | 88.6% | 86.2% |
| MMLU | 69.9% | 66.3% |
| GPQA Diamond | 26.3% | — |
| LMArena Expert | 1039 | — |
| OpenBookQA | 60.2% | — |
| TriviaQA | 87.6% | — |
Multilingual Too close to call
Llama 2-70B: 27.7 (#274), Qwen-14B: 27.5 (#275)
| Benchmark | Llama 2-70B | Qwen-14B |
|---|---|---|
| LMArena Non-English | 1045 | 1041 |
| LMArena Chinese | 995 | 1077 |
| LMArena French | 1090 | — |
| LMArena German | 1041 | — |
| LMArena Japanese | 927 | — |
| LMArena Korean | 964 | — |
| LMArena Russian | 1083 | — |
| LMArena Spanish | 1143 | — |
Instruction Following Llama 2-70B leads
Llama 2-70B: 54.9 (#278), Qwen-14B: 52.4 (#289)
| Benchmark | Llama 2-70B | Qwen-14B |
|---|---|---|
| LMArena Instruction Following | 1071 | 1031 |
Long Context Too close to call
Llama 2-70B: 32.3 (#270), Qwen-14B: 31.3 (#280)
| Benchmark | Llama 2-70B | Qwen-14B |
|---|---|---|
| LMArena Longer Query | 1062 | 1028 |
Writing & Preference Llama 2-70B leads
Llama 2-70B: 32.3 (#279), Qwen-14B: 27.6 (#299)
| Benchmark | Llama 2-70B | Qwen-14B |
|---|---|---|
| LMArena Text | 1115 | 1051 |
| LMArena Creative Writing | 1075 | 1028 |
| LMArena Multi-Turn | 1088 | 1022 |
Frequently asked questions
Is Llama 2-70B better than Qwen-14B?
Qwen-14B is the stronger model overall, scoring 31.4 to 24.4 on the Noometry Index.
Is Llama 2-70B or Qwen-14B better for coding?
They score almost the same on coding (31.4 vs 31.2); test both on your own repository before choosing.
How many benchmarks do Llama 2-70B and Qwen-14B share?
18 benchmarks have published results for both models. Llama 2-70B has 35 scored results on Noometry and Qwen-14B has 18.