Model comparison
Llama 2-7B vs Phi-2
Llama 2-7B has enough public results to be ranked (#317); Phi-2 does not yet, so treat this comparison as directional.
Last verified . 8 shared benchmarks.
Meta
MicrosoftSummary
- They share 8 benchmarks with published results for both.
Side by side
| Llama 2-7B | Phi-2 | |
|---|---|---|
| Provider | Meta | Microsoft |
| Noometry Index | 29.1 | — |
| Released | 2023-07-18 | 2023-12-12 |
| Weights | Open | Open |
| Context window | — | — |
| Max output | — | — |
| Input $ / M tokens | — | — |
| Output $ / M tokens | — | — |
| Results tracked | 29 | 11 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding Not comparable
Llama 2-7B: 29.2 (#307), Phi-2: —
| Benchmark | Llama 2-7B | Phi-2 |
|---|---|---|
| LMArena Coding | 1002 | — |
| HumanEval+ | — | 45.1% |
| MBPP+ | — | 54.2% |
Reasoning Not comparable
Llama 2-7B: 15.7 (#312), Phi-2: —
| Benchmark | Llama 2-7B | Phi-2 |
|---|---|---|
| BIG-Bench Hard | 39.2% | 59.4% |
| Epoch Capabilities Index | 99.06 | 107.94 |
| HellaSwag | 77.2% | 53.6% |
| WinoGrande | 69.2% | 54.7% |
| Chess Puzzles | 0% | — |
| LMArena Hard Prompts | 1009 | — |
| Adversarial NLI | — | 42.5% |
| LAMBADA | 73.3% | — |
| PIQA | 78.8% | — |
Math Not comparable
Llama 2-7B: 30.7 (#233), Phi-2: —
| Benchmark | Llama 2-7B | Phi-2 |
|---|---|---|
| LMArena Math | 1042 | — |
| GSM8K | 16.7% | — |
Knowledge Not comparable
Llama 2-7B: 28.2 (#248), Phi-2: —
| Benchmark | Llama 2-7B | Phi-2 |
|---|---|---|
| ARC (AI2) Challenge | 45.9% | 75.9% |
| MMLU | 45.8% | 58.4% |
| OpenBookQA | 58.6% | 73.6% |
| TriviaQA | 73.7% | 45.2% |
| LMArena Expert | 1036 | — |
| BoolQ | 77.9% | — |
Multimodal Not comparable
Llama 2-7B: —, Phi-2: —
| Benchmark | Llama 2-7B | Phi-2 |
|---|---|---|
| ScienceQA | 43.1% | — |
Multilingual Not comparable
Llama 2-7B: 23.8 (#293), Phi-2: —
| Benchmark | Llama 2-7B | Phi-2 |
|---|---|---|
| LMArena Non-English | 973 | — |
| LMArena Chinese | 973 | — |
| LMArena French | 970 | — |
| LMArena German | 978 | — |
| LMArena Russian | 995 | — |
| LMArena Spanish | 1007 | — |
Instruction Following Not comparable
Llama 2-7B: 50.8 (#298), Phi-2: —
| Benchmark | Llama 2-7B | Phi-2 |
|---|---|---|
| LMArena Instruction Following | 1006 | — |
Long Context Not comparable
Llama 2-7B: 30.4 (#287), Phi-2: —
| Benchmark | Llama 2-7B | Phi-2 |
|---|---|---|
| LMArena Longer Query | 999 | — |
Writing & Preference Not comparable
Llama 2-7B: 28.0 (#298), Phi-2: —
| Benchmark | Llama 2-7B | Phi-2 |
|---|---|---|
| LMArena Text | 1053 | — |
| LMArena Creative Writing | 1033 | — |
| LMArena Multi-Turn | 1029 | — |
Frequently asked questions
Is Llama 2-7B better than Phi-2?
Llama 2-7B has enough public results to be ranked (#317); Phi-2 does not yet, so treat this comparison as directional.
How many benchmarks do Llama 2-7B and Phi-2 share?
8 benchmarks have published results for both models. Llama 2-7B has 29 scored results on Noometry and Phi-2 has 11.