# Llama 2-13B vs Phi-4

> Phi-4 is the stronger model overall, scoring 31.2 to 29.6 on the Noometry Index.

- Canonical page: https://noometry.com/compare/llama-2-13b-vs-phi-4
- Last updated: 2026-10-12
- Shared benchmarks: 20

## Summary

- They share 20 benchmarks with published results for both. Llama 2-13B 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 26.5.

## Snapshot

| | Llama 2-13B | Phi-4 |
|---|---|---|
| Provider | Meta | Microsoft |
| Noometry Index | 29.6 | 31.2 |
| Rank | 309 | 279 |
| Context | — | 128K |
| Input $/M | — | $0.07 |
| Output $/M | — | $0.14 |
| Weights | Open | Open |

## Coding

- Llama 2-13B: 30.9 (#291)
- Phi-4: 34.4 (#239)

| Benchmark | Llama 2-13B | Phi-4 |
|---|---|---|
| LMArena Coding | 1062 | 1231 |
| BigCodeBench Instruct | — | 45.5% |
| LiveBench Coding | — | 30.7% |
| BigCodeBench Complete | — | 55.4% |

## Agentic & Tool Use

- Llama 2-13B: —
- Phi-4: 22.8 (#128)

| Benchmark | Llama 2-13B | Phi-4 |
|---|---|---|
| Berkeley Function Calling Leaderboard | — | 28.8% |
| BALROG | — | 11.6% |

## Reasoning

- Llama 2-13B: 12.8 (#337)
- Phi-4: 17.7 (#291)

| Benchmark | Llama 2-13B | Phi-4 |
|---|---|---|
| Chess Puzzles | 0% | 1% |
| LMArena Hard Prompts | 1051 | 1220 |
| Epoch Capabilities Index | 106.17 | 130.42 |
| LiveBench Reasoning | — | 47.8% |
| DTBench | 42.2% | — |
| LiveBench Data Analysis | — | 45.2% |
| BIG-Bench Hard | 58.2% | — |
| HellaSwag | 80.7% | — |
| LAMBADA | 76.5% | — |
| LiveBench | — | 41.6% |
| PIQA | 80.8% | — |
| WinoGrande | 72.8% | — |

## Math

- Llama 2-13B: 31.1 (#229)
- Phi-4: 20.8 (#285)

| Benchmark | Llama 2-13B | Phi-4 |
|---|---|---|
| LMArena Math | 1065 | 1246 |
| OTIS Mock AIME 2024-2025 | — | 13.8% |
| LiveBench Math | — | 42% |
| MATH Level 5 | — | 64.9% |
| GSM8K | 36.9% | — |

## Knowledge

- Llama 2-13B: 28.1 (#249)
- Phi-4: 32.6 (#209)

| Benchmark | Llama 2-13B | Phi-4 |
|---|---|---|
| LMArena Expert | 1030 | 1203 |
| MMLU | 55.6% | 84.8% |
| GPQA Diamond | — | 56.1% |
| Confabulations | — | 29.4% |
| Vectara Hallucination Rate | — | 3.7% |
| ARC (AI2) Challenge | 60.3% | — |
| BoolQ | 82.4% | — |
| OpenBookQA | 57% | — |
| TriviaQA | 79.6% | — |

## Multimodal

- Llama 2-13B: —
- Phi-4: —

| Benchmark | Llama 2-13B | Phi-4 |
|---|---|---|
| ScienceQA | 55.8% | — |

## Multilingual

- Llama 2-13B: 26.5 (#279)
- Phi-4: 37.2 (#237)

| Benchmark | Llama 2-13B | Phi-4 |
|---|---|---|
| LMArena Non-English | 1024 | 1197 |
| LMArena Chinese | 1001 | 1212 |
| LMArena French | 1044 | 1224 |
| LMArena German | 1009 | 1222 |
| LMArena Japanese | 894 | 1158 |
| LMArena Korean | 953 | 1151 |
| LMArena Russian | 1055 | 1209 |
| LMArena Spanish | 1087 | 1234 |

## Instruction Following

- Llama 2-13B: 53.3 (#287)
- Phi-4: 60.4 (#251)

| Benchmark | Llama 2-13B | Phi-4 |
|---|---|---|
| LMArena Instruction Following | 1045 | 1201 |
| LiveBench Instruction Following | — | 58.4% |

## Long Context

- Llama 2-13B: 32.3 (#269)
- Phi-4: 36.9 (#226)

| Benchmark | Llama 2-13B | Phi-4 |
|---|---|---|
| LMArena Longer Query | 1064 | 1217 |

## Writing & Preference

- Llama 2-13B: 29.8 (#289)
- Phi-4: 40.5 (#244)

| Benchmark | Llama 2-13B | Phi-4 |
|---|---|---|
| LMArena Text | 1084 | 1217 |
| LMArena Creative Writing | 1047 | 1182 |
| LMArena Multi-Turn | 1050 | 1206 |
| Short-Story Creative Writing | — | 62.6% |
| LiveBench Language | — | 25.6% |

## FAQ

### Is Llama 2-13B better than Phi-4?

Phi-4 is the stronger model overall, scoring 31.2 to 29.6 on the Noometry Index.

### Is Llama 2-13B or Phi-4 better for coding?

Phi-4 scores higher on coding benchmarks: 34.4 versus 30.9 in the Noometry coding category.

### How many benchmarks do Llama 2-13B and Phi-4 share?

20 benchmarks have published results for both models. Llama 2-13B has 32 scored results on Noometry and Phi-4 has 37.