Model comparison
Mixtral 8x7B vs Phi 3 Small 8k Instruct
Phi 3 Small 8k Instruct is the stronger model overall, scoring 29.3 to 27.1 on the Noometry Index.
Last verified . 24 shared benchmarks.
Mistral AI
MicrosoftSummary
- They share 24 benchmarks with published results for both. Mixtral 8x7B scores higher in 5 categories and Phi 3 Small 8k Instruct in 3 categories; 6 gaps are clear of the uncertainty.
- The widest gap is in knowledge, where Phi 3 Small 8k Instruct leads 29.1 to 11.0.
Side by side
| Mixtral 8x7B | Phi 3 Small 8k Instruct | |
|---|---|---|
| Provider | Mistral AI | Microsoft |
| Noometry Index | 27.1 | 29.3 |
| Released | 2023-12-11 | 2024-04-23 |
| Weights | Open | Open |
| Context window | 32K | — |
| Max output | 32K | — |
| Input $ / M tokens | $0.70 | — |
| Output $ / M tokens | $0.70 | — |
| Results tracked | 38 | 32 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding Mixtral 8x7B leads
Mixtral 8x7B: 32.8 (#269), Phi 3 Small 8k Instruct: 27.9 (#318)
| Benchmark | Mixtral 8x7B | Phi 3 Small 8k Instruct |
|---|---|---|
| LMArena Coding | 1126 | 1101 |
| LiveBench Coding | — | 20.3% |
| HumanEval+ | 39.6% | — |
| MBPP+ | 49.7% | — |
Reasoning Mixtral 8x7B leads
Mixtral 8x7B: 18.2 (#285), Phi 3 Small 8k Instruct: 14.8 (#323)
| Benchmark | Mixtral 8x7B | Phi 3 Small 8k Instruct |
|---|---|---|
| LMArena Hard Prompts | 1115 | 1100 |
| Adversarial NLI | 55.2% | 58.1% |
| HellaSwag | 86.7% | 77% |
| WinoGrande | 77.2% | 81.5% |
| LiveBench Reasoning | — | 15.9% |
| DTBench | 49.6% | — |
| LiveBench Data Analysis | — | 30.3% |
| BIG-Bench Hard | — | 79.1% |
| Epoch Capabilities Index | 118.47 | — |
| ForecastBench | 56.3 | — |
| LiveBench | — | 24% |
| PIQA | 83.6% | — |
Math Phi 3 Small 8k Instruct leads
Mixtral 8x7B: 18.8 (#289), Phi 3 Small 8k Instruct: 27.6 (#248)
| Benchmark | Mixtral 8x7B | Phi 3 Small 8k Instruct |
|---|---|---|
| LMArena Math | 1147 | 1151 |
| Omni-MATH | 10.5% | — |
| LiveBench Math | — | 17.6% |
| MATH Level 5 | 10% | — |
| GSM8K | 74.4% | — |
Knowledge Phi 3 Small 8k Instruct leads
Mixtral 8x7B: 11.0 (#301), Phi 3 Small 8k Instruct: 29.1 (#240)
| Benchmark | Mixtral 8x7B | Phi 3 Small 8k Instruct |
|---|---|---|
| LMArena Expert | 1088 | 1067 |
| ARC (AI2) Challenge | 87.3% | 90.7% |
| MMLU | 70.6% | 75.7% |
| OpenBookQA | 85.8% | 88% |
| TriviaQA | 82.2% | 58.1% |
| GPQA Diamond | 30.6% | — |
| MMLU-Pro | 33.5% | — |
| GPQA (HELM) | 29.6% | — |
Multilingual Mixtral 8x7B leads
Mixtral 8x7B: 29.6 (#266), Phi 3 Small 8k Instruct: 28.5 (#272)
| Benchmark | Mixtral 8x7B | Phi 3 Small 8k Instruct |
|---|---|---|
| LMArena Non-English | 1077 | 1058 |
| LMArena Chinese | 1055 | 1061 |
| LMArena French | 1166 | 1135 |
| LMArena German | 1114 | 1080 |
| LMArena Japanese | 931 | 966 |
| LMArena Korean | 968 | 894 |
| LMArena Russian | 1090 | 1111 |
| LMArena Spanish | 1111 | 1111 |
Instruction Following Too close to call
Mixtral 8x7B: 51.0 (#297), Phi 3 Small 8k Instruct: 51.9 (#292)
| Benchmark | Mixtral 8x7B | Phi 3 Small 8k Instruct |
|---|---|---|
| LMArena Instruction Following | 1109 | 1087 |
| LiveBench Instruction Following | — | 47.2% |
| IFEval | 57.5% | — |
Long Context Too close to call
Mixtral 8x7B: 33.4 (#260), Phi 3 Small 8k Instruct: 33.0 (#267)
| Benchmark | Mixtral 8x7B | Phi 3 Small 8k Instruct |
|---|---|---|
| LMArena Longer Query | 1103 | 1088 |
Writing & Preference Mixtral 8x7B leads
Mixtral 8x7B: 34.2 (#270), Phi 3 Small 8k Instruct: 31.1 (#284)
| Benchmark | Mixtral 8x7B | Phi 3 Small 8k Instruct |
|---|---|---|
| LMArena Text | 1132 | 1110 |
| LMArena Creative Writing | 1109 | 1083 |
| LMArena Multi-Turn | 1115 | 1068 |
| WildBench | 67.3% | — |
| LiveBench Language | — | 12.9% |
Frequently asked questions
Is Mixtral 8x7B better than Phi 3 Small 8k Instruct?
Phi 3 Small 8k Instruct is the stronger model overall, scoring 29.3 to 27.1 on the Noometry Index.
Is Mixtral 8x7B or Phi 3 Small 8k Instruct better for coding?
Mixtral 8x7B scores higher on coding benchmarks: 32.8 versus 27.9 in the Noometry coding category.
How many benchmarks do Mixtral 8x7B and Phi 3 Small 8k Instruct share?
24 benchmarks have published results for both models. Mixtral 8x7B has 38 scored results on Noometry and Phi 3 Small 8k Instruct has 32.