# Phi 3 Mini 4k Instruct vs Qwen2.5-Coder-32B

> Qwen2.5-Coder-32B is the stronger model overall, scoring 33.4 to 27.9 on the Noometry Index.

- Canonical page: https://noometry.com/compare/phi-3-mini-4k-instruct-vs-qwen2-5-coder-32b
- Last updated: 2026-10-11
- Shared benchmarks: 25

## Summary

- They share 25 benchmarks with published results for both. Phi 3 Mini 4k Instruct scores higher in 1 category and Qwen2.5-Coder-32B in 7 categories; 8 gaps are clear of the uncertainty.
- The widest gap is in writing & preference, where Qwen2.5-Coder-32B leads 41.6 to 27.6.
- The biggest single-benchmark swing is LiveBench Coding: 15.5% for Phi 3 Mini 4k Instruct and 56.9% for Qwen2.5-Coder-32B.

## Snapshot

| | Phi 3 Mini 4k Instruct | Qwen2.5-Coder-32B |
|---|---|---|
| Provider | Microsoft | Alibaba (Qwen) |
| Noometry Index | 27.9 | 33.4 |
| Rank | 328 | 245 |
| Context | — | 33K |
| Input $/M | — | $0.66 |
| Output $/M | — | $1 |
| Weights | Open | Open |

## Coding

- Phi 3 Mini 4k Instruct: 26.6 (#323)
- Qwen2.5-Coder-32B: 22.6 (#333)

| Benchmark | Phi 3 Mini 4k Instruct | Qwen2.5-Coder-32B |
|---|---|---|
| LiveBench Coding | 15.5% | 56.9% |
| LMArena Coding | 1093 | 1276 |
| HumanEval+ | 59.1% | 87.2% |
| MBPP+ | 54.2% | 77% |
| SWE-bench Verified (bash only) | — | 9% |
| Aider Polyglot | — | 16.4% |
| BigCodeBench Instruct | — | 49% |
| BigCodeBench Complete | — | 58% |

## Reasoning

- Phi 3 Mini 4k Instruct: 14.1 (#328)
- Qwen2.5-Coder-32B: 21.2 (#225)

| Benchmark | Phi 3 Mini 4k Instruct | Qwen2.5-Coder-32B |
|---|---|---|
| LiveBench Reasoning | 26.8% | 42.1% |
| LMArena Hard Prompts | 1072 | 1251 |
| LiveBench Data Analysis | 34.7% | 49.9% |
| HellaSwag | 76.7% | 83% |
| LiveBench | 22.4% | 46.2% |
| WinoGrande | 70.8% | 80.8% |
| Chess Puzzles | 0% | — |
| Adversarial NLI | 52.8% | — |
| BIG-Bench Hard | 71.7% | — |
| Epoch Capabilities Index | — | 119.49 |

## Math

- Phi 3 Mini 4k Instruct: 26.6 (#257)
- Qwen2.5-Coder-32B: 33.3 (#204)

| Benchmark | Phi 3 Mini 4k Instruct | Qwen2.5-Coder-32B |
|---|---|---|
| LiveBench Math | 15.7% | 46.6% |
| LMArena Math | 1111 | 1251 |
| GSM8K | — | 93% |

## Knowledge

- Phi 3 Mini 4k Instruct: 28.5 (#246)
- Qwen2.5-Coder-32B: 33.4 (#203)

| Benchmark | Phi 3 Mini 4k Instruct | Qwen2.5-Coder-32B |
|---|---|---|
| LMArena Expert | 1045 | 1221 |
| ARC (AI2) Challenge | 84.9% | 70.5% |
| MMLU | 68.8% | 79.1% |
| OpenBookQA | 88% | — |
| TriviaQA | 64% | — |

## Multilingual

- Phi 3 Mini 4k Instruct: 26.3 (#280)
- Qwen2.5-Coder-32B: 37.8 (#235)

| Benchmark | Phi 3 Mini 4k Instruct | Qwen2.5-Coder-32B |
|---|---|---|
| LMArena Non-English | 1021 | 1205 |
| LMArena Chinese | 1021 | 1222 |
| LMArena Russian | 1022 | 1228 |
| LMArena French | 1076 | — |
| LMArena German | 1044 | — |
| LMArena Japanese | 935 | — |
| LMArena Korean | 905 | — |
| LMArena Spanish | 1085 | — |

## Instruction Following

- Phi 3 Mini 4k Instruct: 47.7 (#303)
- Qwen2.5-Coder-32B: 61.4 (#245)

| Benchmark | Phi 3 Mini 4k Instruct | Qwen2.5-Coder-32B |
|---|---|---|
| LiveBench Instruction Following | 39.1% | 58.7% |
| LMArena Instruction Following | 1053 | 1223 |

## Long Context

- Phi 3 Mini 4k Instruct: 31.7 (#276)
- Qwen2.5-Coder-32B: 38.0 (#208)

| Benchmark | Phi 3 Mini 4k Instruct | Qwen2.5-Coder-32B |
|---|---|---|
| LMArena Longer Query | 1044 | 1251 |

## Writing & Preference

- Phi 3 Mini 4k Instruct: 27.6 (#300)
- Qwen2.5-Coder-32B: 41.6 (#240)

| Benchmark | Phi 3 Mini 4k Instruct | Qwen2.5-Coder-32B |
|---|---|---|
| LMArena Text | 1073 | 1230 |
| LMArena Creative Writing | 1037 | 1174 |
| LMArena Multi-Turn | 1018 | 1222 |
| LiveBench Language | 9.2% | 23.3% |

## FAQ

### Is Phi 3 Mini 4k Instruct better than Qwen2.5-Coder-32B?

Qwen2.5-Coder-32B is the stronger model overall, scoring 33.4 to 27.9 on the Noometry Index.

### Is Phi 3 Mini 4k Instruct or Qwen2.5-Coder-32B better for coding?

Phi 3 Mini 4k Instruct scores higher on coding benchmarks: 26.6 versus 22.6 in the Noometry coding category.

### How many benchmarks do Phi 3 Mini 4k Instruct and Qwen2.5-Coder-32B share?

25 benchmarks have published results for both models. Phi 3 Mini 4k Instruct has 35 scored results on Noometry and Qwen2.5-Coder-32B has 31.