Model comparison
Mixtral 8x7B vs Qwen-14B
Qwen-14B is the stronger model overall, scoring 31.4 to 27.1 on the Noometry Index.
Last verified . 15 shared benchmarks.
Mistral AI
Alibaba (Qwen)Summary
- They share 15 benchmarks with published results for both. Mixtral 8x7B scores higher in 4 categories and Qwen-14B in 3 categories; 7 gaps are clear of the uncertainty.
- The widest gap is in math, where Qwen-14B leads 31.2 to 18.8.
Side by side
| Mixtral 8x7B | Qwen-14B | |
|---|---|---|
| Provider | Mistral AI | Alibaba (Qwen) |
| Noometry Index | 27.1 | 31.4 |
| Released | 2023-12-11 | 2023-09-24 |
| Weights | Open | Open |
| Context window | 32K | — |
| Max output | 32K | — |
| Input $ / M tokens | $0.70 | — |
| Output $ / M tokens | $0.70 | — |
| Results tracked | 38 | 18 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding Mixtral 8x7B leads
Mixtral 8x7B: 32.8 (#269), Qwen-14B: 31.2 (#288)
| Benchmark | Mixtral 8x7B | Qwen-14B |
|---|---|---|
| LMArena Coding | 1126 | 1071 |
| HumanEval+ | 39.6% | — |
| MBPP+ | 49.7% | — |
Reasoning Qwen-14B leads
Mixtral 8x7B: 18.2 (#285), Qwen-14B: 19.6 (#257)
| Benchmark | Mixtral 8x7B | Qwen-14B |
|---|---|---|
| LMArena Hard Prompts | 1115 | 1027 |
| Epoch Capabilities Index | 118.47 | 113.03 |
| PIQA | 83.6% | 79.9% |
| DTBench | 49.6% | — |
| Adversarial NLI | 55.2% | — |
| BIG-Bench Hard | — | 55% |
| ForecastBench | 56.3 | — |
| HellaSwag | 86.7% | — |
| LAMBADA | — | 71.1% |
| WinoGrande | 77.2% | — |
Math Qwen-14B leads
Mixtral 8x7B: 18.8 (#289), Qwen-14B: 31.2 (#227)
| Benchmark | Mixtral 8x7B | Qwen-14B |
|---|---|---|
| LMArena Math | 1147 | 1068 |
| GSM8K | 74.4% | 61.3% |
| Omni-MATH | 10.5% | — |
| MATH Level 5 | 10% | — |
Knowledge Not comparable
Mixtral 8x7B: 11.0 (#301), Qwen-14B: —
| Benchmark | Mixtral 8x7B | Qwen-14B |
|---|---|---|
| ARC (AI2) Challenge | 87.3% | 84.4% |
| MMLU | 70.6% | 66.3% |
| GPQA Diamond | 30.6% | — |
| MMLU-Pro | 33.5% | — |
| GPQA (HELM) | 29.6% | — |
| LMArena Expert | 1088 | — |
| BoolQ | — | 86.2% |
| OpenBookQA | 85.8% | — |
| TriviaQA | 82.2% | — |
Multilingual Mixtral 8x7B leads
Mixtral 8x7B: 29.6 (#266), Qwen-14B: 27.5 (#275)
| Benchmark | Mixtral 8x7B | Qwen-14B |
|---|---|---|
| LMArena Non-English | 1077 | 1041 |
| LMArena Chinese | 1055 | 1077 |
| LMArena French | 1166 | — |
| LMArena German | 1114 | — |
| LMArena Japanese | 931 | — |
| LMArena Korean | 968 | — |
| LMArena Russian | 1090 | — |
| LMArena Spanish | 1111 | — |
Instruction Following Qwen-14B leads
Mixtral 8x7B: 51.0 (#297), Qwen-14B: 52.4 (#289)
| Benchmark | Mixtral 8x7B | Qwen-14B |
|---|---|---|
| LMArena Instruction Following | 1109 | 1031 |
| IFEval | 57.5% | — |
Long Context Mixtral 8x7B leads
Mixtral 8x7B: 33.4 (#260), Qwen-14B: 31.3 (#280)
| Benchmark | Mixtral 8x7B | Qwen-14B |
|---|---|---|
| LMArena Longer Query | 1103 | 1028 |
Writing & Preference Mixtral 8x7B leads
Mixtral 8x7B: 34.2 (#270), Qwen-14B: 27.6 (#299)
| Benchmark | Mixtral 8x7B | Qwen-14B |
|---|---|---|
| LMArena Text | 1132 | 1051 |
| LMArena Creative Writing | 1109 | 1028 |
| LMArena Multi-Turn | 1115 | 1022 |
| WildBench | 67.3% | — |
Frequently asked questions
Is Mixtral 8x7B better than Qwen-14B?
Qwen-14B is the stronger model overall, scoring 31.4 to 27.1 on the Noometry Index.
Is Mixtral 8x7B or Qwen-14B better for coding?
Mixtral 8x7B scores higher on coding benchmarks: 32.8 versus 31.2 in the Noometry coding category.
How many benchmarks do Mixtral 8x7B and Qwen-14B share?
15 benchmarks have published results for both models. Mixtral 8x7B has 38 scored results on Noometry and Qwen-14B has 18.