# phi-3-medium 14B vs Qwen3 32B

> Qwen3 32B is the stronger model overall, scoring 38.8 to 29.1 on the Noometry Index.

- Canonical page: https://noometry.com/compare/phi-3-medium-14b-vs-qwen3-32b
- Last updated: 2026-10-10
- Shared benchmarks: 2

## Summary

- They share 2 benchmarks with published results for both. phi-3-medium 14B scores higher in 0 categories and Qwen3 32B in 3 categories; 3 gaps are clear of the uncertainty.
- The widest gap is in knowledge, where Qwen3 32B leads 40.0 to 9.1.
- The biggest single-benchmark swing is GPQA Diamond: 27.6% for phi-3-medium 14B and 65.7% for Qwen3 32B.

## Snapshot

| | phi-3-medium 14B | Qwen3 32B |
|---|---|---|
| Provider | Microsoft | Alibaba (Qwen) |
| Noometry Index | 29.1 | 38.8 |
| Rank | 306 | 166 |
| Context | — | 131K |
| Input $/M | — | $0.70 |
| Output $/M | — | $2.80 |
| Weights | Open | Open |

## Coding

- phi-3-medium 14B: 36.3 (#198)
- Qwen3 32B: 37.3 (#184)

| Benchmark | phi-3-medium 14B | Qwen3 32B |
|---|---|---|
| Aider Polyglot | — | 40% |
| SciCode | — | 35.4% |
| BigCodeBench Instruct | 37.6% | — |
| LMArena Coding | — | 1358 |
| BigCodeBench Complete | 48.7% | — |

## Agentic & Tool Use

- phi-3-medium 14B: —
- Qwen3 32B: 32.3 (#58)

| Benchmark | phi-3-medium 14B | Qwen3 32B |
|---|---|---|
| Berkeley Function Calling Leaderboard | — | 48.7% |

## Reasoning

- phi-3-medium 14B: —
- Qwen3 32B: 19.6 (#233)

| Benchmark | phi-3-medium 14B | Qwen3 32B |
|---|---|---|
| Epoch Capabilities Index | 121.18 | 138.51 |
| Kagi LLM Benchmark | — | 54.9% |
| CritPt | — | 0.3% |
| Chess Puzzles | — | 5% |
| LMArena Hard Prompts | — | 1335 |
| DTBench | — | 67.5% |
| LMCA | — | 17.3% |
| Adversarial NLI | 55.8% | — |
| BIG-Bench Hard | 81.4% | — |
| HellaSwag | 82.4% | — |
| WinoGrande | 81.5% | — |

## Math

- phi-3-medium 14B: 25.2 (#249)
- Qwen3 32B: 37.9 (#89)

| Benchmark | phi-3-medium 14B | Qwen3 32B |
|---|---|---|
| OTIS Mock AIME 2024-2025 | — | 66.9% |
| LMArena Math | — | 1399 |
| MATH Level 5 | 17.6% | — |

## Knowledge

- phi-3-medium 14B: 9.1 (#301)
- Qwen3 32B: 40.0 (#112)

| Benchmark | phi-3-medium 14B | Qwen3 32B |
|---|---|---|
| GPQA Diamond | 27.6% | 65.7% |
| Vectara Hallucination Rate | — | 5.9% |
| LMArena Expert | — | 1362 |
| ARC (AI2) Challenge | 91.6% | — |
| MMLU | 78% | — |
| OpenBookQA | 87.4% | — |
| TriviaQA | 73.9% | — |

## Multilingual

- phi-3-medium 14B: —
- Qwen3 32B: 45.8 (#162)

| Benchmark | phi-3-medium 14B | Qwen3 32B |
|---|---|---|
| LMArena Non-English | — | 1317 |
| LMArena Chinese | — | 1357 |
| LMArena German | — | 1341 |
| LMArena Russian | — | 1311 |

## Instruction Following

- phi-3-medium 14B: —
- Qwen3 32B: 69.1 (#174)

| Benchmark | phi-3-medium 14B | Qwen3 32B |
|---|---|---|
| LMArena Instruction Following | — | 1305 |

## Long Context

- phi-3-medium 14B: —
- Qwen3 32B: 43.7 (#83)

| Benchmark | phi-3-medium 14B | Qwen3 32B |
|---|---|---|
| Fiction.LiveBench | — | 74.2% |
| LMArena Longer Query | — | 1327 |

## Writing & Preference

- phi-3-medium 14B: —
- Qwen3 32B: 53.1 (#158)

| Benchmark | phi-3-medium 14B | Qwen3 32B |
|---|---|---|
| LMArena Text | — | 1340 |
| LMArena Creative Writing | — | 1297 |
| LMArena Multi-Turn | — | 1331 |

## FAQ

### Is phi-3-medium 14B better than Qwen3 32B?

Qwen3 32B is the stronger model overall, scoring 38.8 to 29.1 on the Noometry Index.

### Is phi-3-medium 14B or Qwen3 32B better for coding?

Qwen3 32B scores higher on coding benchmarks: 37.3 versus 36.3 in the Noometry coding category.

### How many benchmarks do phi-3-medium 14B and Qwen3 32B share?

2 benchmarks have published results for both models. phi-3-medium 14B has 13 scored results on Noometry and Qwen3 32B has 26.