Model comparison
Llama 2-7B vs Phi-4
Phi-4 is the stronger model overall, scoring 31.2 to 29.1 on the Noometry Index.
Last verified . 18 shared benchmarks.
Meta
MicrosoftSummary
- They share 18 benchmarks with published results for both. Llama 2-7B scores higher in 1 category and Phi-4 in 7 categories; 8 gaps are clear of the uncertainty.
- The widest gap is in multilingual, where Phi-4 leads 37.2 to 23.8.
Side by side
| Llama 2-7B | Phi-4 | |
|---|---|---|
| Provider | Meta | Microsoft |
| Noometry Index | 29.1 | 31.2 |
| Released | 2023-07-18 | 2024-12-11 |
| Weights | Open | Open |
| Context window | — | 128K |
| Max output | — | 4K |
| Input $ / M tokens | — | $0.07 |
| Output $ / M tokens | — | $0.14 |
| Results tracked | 29 | 37 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding Phi-4 leads
Llama 2-7B: 29.2 (#307), Phi-4: 34.4 (#239)
| Benchmark | Llama 2-7B | Phi-4 |
|---|---|---|
| LMArena Coding | 1002 | 1231 |
| BigCodeBench Instruct | — | 45.5% |
| LiveBench Coding | — | 30.7% |
| BigCodeBench Complete | — | 55.4% |
Agentic & Tool Use Not comparable
Llama 2-7B: —, Phi-4: 22.8 (#128)
| Benchmark | Llama 2-7B | Phi-4 |
|---|---|---|
| Berkeley Function Calling Leaderboard | — | 28.8% |
| BALROG | — | 11.6% |
Reasoning Phi-4 leads
Llama 2-7B: 15.7 (#312), Phi-4: 17.7 (#291)
| Benchmark | Llama 2-7B | Phi-4 |
|---|---|---|
| Chess Puzzles | 0% | 1% |
| LMArena Hard Prompts | 1009 | 1220 |
| Epoch Capabilities Index | 99.06 | 130.42 |
| LiveBench Reasoning | — | 47.8% |
| LiveBench Data Analysis | — | 45.2% |
| BIG-Bench Hard | 39.2% | — |
| HellaSwag | 77.2% | — |
| LAMBADA | 73.3% | — |
| LiveBench | — | 41.6% |
| PIQA | 78.8% | — |
| WinoGrande | 69.2% | — |
Math Llama 2-7B leads
Llama 2-7B: 30.7 (#233), Phi-4: 20.8 (#285)
| Benchmark | Llama 2-7B | Phi-4 |
|---|---|---|
| LMArena Math | 1042 | 1246 |
| OTIS Mock AIME 2024-2025 | — | 13.8% |
| LiveBench Math | — | 42% |
| MATH Level 5 | — | 64.9% |
| GSM8K | 16.7% | — |
Knowledge Phi-4 leads
Llama 2-7B: 28.2 (#248), Phi-4: 32.6 (#209)
| Benchmark | Llama 2-7B | Phi-4 |
|---|---|---|
| LMArena Expert | 1036 | 1203 |
| MMLU | 45.8% | 84.8% |
| GPQA Diamond | — | 56.1% |
| Confabulations | — | 29.4% |
| Vectara Hallucination Rate | — | 3.7% |
| ARC (AI2) Challenge | 45.9% | — |
| BoolQ | 77.9% | — |
| OpenBookQA | 58.6% | — |
| TriviaQA | 73.7% | — |
Multimodal Not comparable
Llama 2-7B: —, Phi-4: —
| Benchmark | Llama 2-7B | Phi-4 |
|---|---|---|
| ScienceQA | 43.1% | — |
Multilingual Phi-4 leads
Llama 2-7B: 23.8 (#293), Phi-4: 37.2 (#237)
| Benchmark | Llama 2-7B | Phi-4 |
|---|---|---|
| LMArena Non-English | 973 | 1197 |
| LMArena Chinese | 973 | 1212 |
| LMArena French | 970 | 1224 |
| LMArena German | 978 | 1222 |
| LMArena Russian | 995 | 1209 |
| LMArena Spanish | 1007 | 1234 |
| LMArena Japanese | — | 1158 |
| LMArena Korean | — | 1151 |
Instruction Following Phi-4 leads
Llama 2-7B: 50.8 (#298), Phi-4: 60.4 (#251)
| Benchmark | Llama 2-7B | Phi-4 |
|---|---|---|
| LMArena Instruction Following | 1006 | 1201 |
| LiveBench Instruction Following | — | 58.4% |
Long Context Phi-4 leads
Llama 2-7B: 30.4 (#287), Phi-4: 36.9 (#226)
| Benchmark | Llama 2-7B | Phi-4 |
|---|---|---|
| LMArena Longer Query | 999 | 1217 |
Writing & Preference Phi-4 leads
Llama 2-7B: 28.0 (#298), Phi-4: 40.5 (#244)
| Benchmark | Llama 2-7B | Phi-4 |
|---|---|---|
| LMArena Text | 1053 | 1217 |
| LMArena Creative Writing | 1033 | 1182 |
| LMArena Multi-Turn | 1029 | 1206 |
| Short-Story Creative Writing | — | 62.6% |
| LiveBench Language | — | 25.6% |
Frequently asked questions
Is Llama 2-7B better than Phi-4?
Phi-4 is the stronger model overall, scoring 31.2 to 29.1 on the Noometry Index.
Is Llama 2-7B or Phi-4 better for coding?
Phi-4 scores higher on coding benchmarks: 34.4 versus 29.2 in the Noometry coding category.
How many benchmarks do Llama 2-7B and Phi-4 share?
18 benchmarks have published results for both models. Llama 2-7B has 29 scored results on Noometry and Phi-4 has 37.