Model comparison
Mixtral 8x7B vs Phi-2
Mixtral 8x7B has enough public results to be ranked (#334); Phi-2 does not yet, so treat this comparison as directional.
Last verified . 10 shared benchmarks.
Mistral AI
MicrosoftSummary
- They share 10 benchmarks with published results for both.
Side by side
| Mixtral 8x7B | Phi-2 | |
|---|---|---|
| Provider | Mistral AI | Microsoft |
| Noometry Index | 27.1 | — |
| Released | 2023-12-11 | 2023-12-12 |
| Weights | Open | Open |
| Context window | 32K | — |
| Max output | 32K | — |
| Input $ / M tokens | $0.70 | — |
| Output $ / M tokens | $0.70 | — |
| Results tracked | 38 | 11 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding Not comparable
Mixtral 8x7B: 32.8 (#269), Phi-2: —
| Benchmark | Mixtral 8x7B | Phi-2 |
|---|---|---|
| HumanEval+ | 39.6% | 45.1% |
| MBPP+ | 49.7% | 54.2% |
| LMArena Coding | 1126 | — |
Reasoning Not comparable
Mixtral 8x7B: 18.2 (#285), Phi-2: —
| Benchmark | Mixtral 8x7B | Phi-2 |
|---|---|---|
| Adversarial NLI | 55.2% | 42.5% |
| Epoch Capabilities Index | 118.47 | 107.94 |
| HellaSwag | 86.7% | 53.6% |
| WinoGrande | 77.2% | 54.7% |
| LMArena Hard Prompts | 1115 | — |
| DTBench | 49.6% | — |
| BIG-Bench Hard | — | 59.4% |
| ForecastBench | 56.3 | — |
| PIQA | 83.6% | — |
Math Not comparable
Mixtral 8x7B: 18.8 (#289), Phi-2: —
| Benchmark | Mixtral 8x7B | Phi-2 |
|---|---|---|
| Omni-MATH | 10.5% | — |
| LMArena Math | 1147 | — |
| MATH Level 5 | 10% | — |
| GSM8K | 74.4% | — |
Knowledge Not comparable
Mixtral 8x7B: 11.0 (#301), Phi-2: —
| Benchmark | Mixtral 8x7B | Phi-2 |
|---|---|---|
| ARC (AI2) Challenge | 87.3% | 75.9% |
| MMLU | 70.6% | 58.4% |
| OpenBookQA | 85.8% | 73.6% |
| TriviaQA | 82.2% | 45.2% |
| GPQA Diamond | 30.6% | — |
| MMLU-Pro | 33.5% | — |
| GPQA (HELM) | 29.6% | — |
| LMArena Expert | 1088 | — |
Multilingual Not comparable
Mixtral 8x7B: 29.6 (#266), Phi-2: —
| Benchmark | Mixtral 8x7B | Phi-2 |
|---|---|---|
| LMArena Non-English | 1077 | — |
| LMArena Chinese | 1055 | — |
| LMArena French | 1166 | — |
| LMArena German | 1114 | — |
| LMArena Japanese | 931 | — |
| LMArena Korean | 968 | — |
| LMArena Russian | 1090 | — |
| LMArena Spanish | 1111 | — |
Instruction Following Not comparable
Mixtral 8x7B: 51.0 (#297), Phi-2: —
| Benchmark | Mixtral 8x7B | Phi-2 |
|---|---|---|
| IFEval | 57.5% | — |
| LMArena Instruction Following | 1109 | — |
Long Context Not comparable
Mixtral 8x7B: 33.4 (#260), Phi-2: —
| Benchmark | Mixtral 8x7B | Phi-2 |
|---|---|---|
| LMArena Longer Query | 1103 | — |
Writing & Preference Not comparable
Mixtral 8x7B: 34.2 (#270), Phi-2: —
| Benchmark | Mixtral 8x7B | Phi-2 |
|---|---|---|
| LMArena Text | 1132 | — |
| LMArena Creative Writing | 1109 | — |
| WildBench | 67.3% | — |
| LMArena Multi-Turn | 1115 | — |
Frequently asked questions
Is Mixtral 8x7B better than Phi-2?
Mixtral 8x7B has enough public results to be ranked (#334); Phi-2 does not yet, so treat this comparison as directional.
How many benchmarks do Mixtral 8x7B and Phi-2 share?
10 benchmarks have published results for both models. Mixtral 8x7B has 38 scored results on Noometry and Phi-2 has 11.