Model comparison
Llama 2-70B vs Phi 3 Small 8k Instruct
Phi 3 Small 8k Instruct is the stronger model overall, scoring 29.3 to 24.4 on the Noometry Index.
Last verified . 24 shared benchmarks.
Meta
MicrosoftSummary
- They share 24 benchmarks with published results for both. Llama 2-70B scores higher in 3 categories and Phi 3 Small 8k Instruct in 5 categories; 5 gaps are clear of the uncertainty.
- The widest gap is in knowledge, where Phi 3 Small 8k Instruct leads 29.1 to 7.4.
Side by side
| Llama 2-70B | Phi 3 Small 8k Instruct | |
|---|---|---|
| Provider | Meta | Microsoft |
| Noometry Index | 24.4 | 29.3 |
| Released | 2023-07-18 | 2024-04-23 |
| Weights | Open | Open |
| Context window | — | — |
| Max output | — | — |
| Input $ / M tokens | — | — |
| Output $ / M tokens | — | — |
| Results tracked | 35 | 32 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding Llama 2-70B leads
Llama 2-70B: 31.4 (#286), Phi 3 Small 8k Instruct: 27.9 (#318)
| Benchmark | Llama 2-70B | Phi 3 Small 8k Instruct |
|---|---|---|
| LMArena Coding | 1079 | 1101 |
| LiveBench Coding | — | 20.3% |
Reasoning Too close to call
Llama 2-70B: 14.4 (#325), Phi 3 Small 8k Instruct: 14.8 (#323)
| Benchmark | Llama 2-70B | Phi 3 Small 8k Instruct |
|---|---|---|
| LMArena Hard Prompts | 1073 | 1100 |
| BIG-Bench Hard | 64.9% | 79.1% |
| HellaSwag | 85.3% | 77% |
| WinoGrande | 80.2% | 81.5% |
| LiveBench Reasoning | — | 15.9% |
| DTBench | 41.6% | — |
| LiveBench Data Analysis | — | 30.3% |
| Adversarial NLI | — | 58.1% |
| CommonsenseQA 2.0 | 50% | — |
| Epoch Capabilities Index | 113.79 | — |
| ForecastBench | 51.4 | — |
| LAMBADA | 78.9% | — |
| LiveBench | — | 24% |
| PIQA | 82.8% | — |
Math Phi 3 Small 8k Instruct leads
Llama 2-70B: 8.1 (#326), Phi 3 Small 8k Instruct: 27.6 (#248)
| Benchmark | Llama 2-70B | Phi 3 Small 8k Instruct |
|---|---|---|
| LMArena Math | 1091 | 1151 |
| OTIS Mock AIME 2024-2025 | 0% | — |
| LiveBench Math | — | 17.6% |
| MATH Level 5 | 3.3% | — |
| GSM8K | 69.6% | — |
Knowledge Phi 3 Small 8k Instruct leads
Llama 2-70B: 7.4 (#310), Phi 3 Small 8k Instruct: 29.1 (#240)
| Benchmark | Llama 2-70B | Phi 3 Small 8k Instruct |
|---|---|---|
| LMArena Expert | 1039 | 1067 |
| ARC (AI2) Challenge | 78.3% | 90.7% |
| MMLU | 69.9% | 75.7% |
| OpenBookQA | 60.2% | 88% |
| TriviaQA | 87.6% | 58.1% |
| GPQA Diamond | 26.3% | — |
| BoolQ | 88.6% | — |
Multilingual Too close to call
Llama 2-70B: 27.7 (#274), Phi 3 Small 8k Instruct: 28.5 (#272)
| Benchmark | Llama 2-70B | Phi 3 Small 8k Instruct |
|---|---|---|
| LMArena Non-English | 1045 | 1058 |
| LMArena Chinese | 995 | 1061 |
| LMArena French | 1090 | 1135 |
| LMArena German | 1041 | 1080 |
| LMArena Japanese | 927 | 966 |
| LMArena Korean | 964 | 894 |
| LMArena Russian | 1083 | 1111 |
| LMArena Spanish | 1143 | 1111 |
Instruction Following Llama 2-70B leads
Llama 2-70B: 54.9 (#278), Phi 3 Small 8k Instruct: 51.9 (#292)
| Benchmark | Llama 2-70B | Phi 3 Small 8k Instruct |
|---|---|---|
| LMArena Instruction Following | 1071 | 1087 |
| LiveBench Instruction Following | — | 47.2% |
Long Context Too close to call
Llama 2-70B: 32.3 (#270), Phi 3 Small 8k Instruct: 33.0 (#267)
| Benchmark | Llama 2-70B | Phi 3 Small 8k Instruct |
|---|---|---|
| LMArena Longer Query | 1062 | 1088 |
Writing & Preference Llama 2-70B leads
Llama 2-70B: 32.3 (#279), Phi 3 Small 8k Instruct: 31.1 (#284)
| Benchmark | Llama 2-70B | Phi 3 Small 8k Instruct |
|---|---|---|
| LMArena Text | 1115 | 1110 |
| LMArena Creative Writing | 1075 | 1083 |
| LMArena Multi-Turn | 1088 | 1068 |
| LiveBench Language | — | 12.9% |
Frequently asked questions
Is Llama 2-70B better than Phi 3 Small 8k Instruct?
Phi 3 Small 8k Instruct is the stronger model overall, scoring 29.3 to 24.4 on the Noometry Index.
Is Llama 2-70B or Phi 3 Small 8k Instruct better for coding?
Llama 2-70B scores higher on coding benchmarks: 31.4 versus 27.9 in the Noometry coding category.
How many benchmarks do Llama 2-70B and Phi 3 Small 8k Instruct share?
24 benchmarks have published results for both models. Llama 2-70B has 35 scored results on Noometry and Phi 3 Small 8k Instruct has 32.