Model comparison
Llama 13b vs Qwen-14B
Qwen-14B is the stronger model overall, scoring 31.4 to 24.4 on the Noometry Index.
Last verified . 16 shared benchmarks.
Meta
Alibaba (Qwen)Summary
- They share 16 benchmarks with published results for both. Llama 13b scores higher in 0 categories and Qwen-14B in 6 categories; 6 gaps are clear of the uncertainty.
- The widest gap is in instruction following, where Qwen-14B leads 52.4 to 36.7.
Side by side
| Llama 13b | Qwen-14B | |
|---|---|---|
| Provider | Meta | Alibaba (Qwen) |
| Noometry Index | 24.4 | 31.4 |
| Released | 2023-02-24 | 2023-09-24 |
| Weights | Open | Open |
| Context window | — | — |
| Max output | — | — |
| Input $ / M tokens | — | — |
| Output $ / M tokens | — | — |
| Results tracked | 21 | 18 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding Qwen-14B leads
Llama 13b: 21.4 (#337), Qwen-14B: 31.2 (#288)
| Benchmark | Llama 13b | Qwen-14B |
|---|---|---|
| LMArena Coding | 683 | 1071 |
Reasoning Qwen-14B leads
Llama 13b: 14.0 (#329), Qwen-14B: 19.6 (#257)
| Benchmark | Llama 13b | Qwen-14B |
|---|---|---|
| LMArena Hard Prompts | 728 | 1027 |
| BIG-Bench Hard | 37.9% | 55% |
| Epoch Capabilities Index | 100.58 | 113.03 |
| LAMBADA | 75.2% | 71.1% |
| PIQA | 80.1% | 79.9% |
| HellaSwag | 79.2% | — |
| WinoGrande | 73% | — |
Math Qwen-14B leads
Llama 13b: 26.7 (#256), Qwen-14B: 31.2 (#227)
| Benchmark | Llama 13b | Qwen-14B |
|---|---|---|
| LMArena Math | 838 | 1068 |
| GSM8K | 20.6% | 61.3% |
Knowledge Not comparable
Llama 13b: —, Qwen-14B: —
| Benchmark | Llama 13b | Qwen-14B |
|---|---|---|
| ARC (AI2) Challenge | 52.7% | 84.4% |
| BoolQ | 78.7% | 86.2% |
| MMLU | 47.7% | 66.3% |
| OpenBookQA | 56.4% | — |
| TriviaQA | 77.9% | — |
Multimodal Not comparable
Llama 13b: —, Qwen-14B: —
| Benchmark | Llama 13b | Qwen-14B |
|---|---|---|
| ScienceQA | 43.3% | — |
Multilingual Qwen-14B leads
Llama 13b: 16.6 (#297), Qwen-14B: 27.5 (#275)
| Benchmark | Llama 13b | Qwen-14B |
|---|---|---|
| LMArena Non-English | 819 | 1041 |
| LMArena Chinese | — | 1077 |
Instruction Following Qwen-14B leads
Llama 13b: 36.7 (#305), Qwen-14B: 52.4 (#289)
| Benchmark | Llama 13b | Qwen-14B |
|---|---|---|
| LMArena Instruction Following | 781 | 1031 |
Long Context Not comparable
Llama 13b: —, Qwen-14B: 31.3 (#280)
| Benchmark | Llama 13b | Qwen-14B |
|---|---|---|
| LMArena Longer Query | — | 1028 |
Writing & Preference Qwen-14B leads
Llama 13b: 13.8 (#312), Qwen-14B: 27.6 (#299)
| Benchmark | Llama 13b | Qwen-14B |
|---|---|---|
| LMArena Text | 834 | 1051 |
| LMArena Creative Writing | 794 | 1028 |
| LMArena Multi-Turn | 753 | 1022 |
Frequently asked questions
Is Llama 13b better than Qwen-14B?
Qwen-14B is the stronger model overall, scoring 31.4 to 24.4 on the Noometry Index.
Is Llama 13b or Qwen-14B better for coding?
Qwen-14B scores higher on coding benchmarks: 31.2 versus 21.4 in the Noometry coding category.
How many benchmarks do Llama 13b and Qwen-14B share?
16 benchmarks have published results for both models. Llama 13b has 21 scored results on Noometry and Qwen-14B has 18.