# Qwen2.5-Coder-32B vs Qwen3 32B

> Qwen3 32B is the stronger model overall, scoring 38.8 to 32.9 on the Noometry Index. Qwen2.5-Coder-32B costs 1.6× less per token, which makes it the better buy when Qwen3 32B's lead doesn't matter for your workload.

- Canonical page: https://noometry.com/compare/qwen2-5-coder-32b-vs-qwen3-32b
- Last updated: 2026-10-10
- Shared benchmarks: 14

## Summary

- They share 14 benchmarks with published results for both. Qwen2.5-Coder-32B scores higher in 1 category and Qwen3 32B in 7 categories; 7 gaps are clear of the uncertainty.
- The widest gap is in coding, where Qwen3 32B leads 37.3 to 22.4.
- The biggest single-benchmark swing is Aider Polyglot: 16.4% for Qwen2.5-Coder-32B and 40% for Qwen3 32B.
- Qwen2.5-Coder-32B is cheaper at $0.66 / $1 per million input/output tokens, against $0.70 / $2.80 for Qwen3 32B.
- Qwen3 32B accepts more context: 131K tokens versus 33K.

## Snapshot

| | Qwen2.5-Coder-32B | Qwen3 32B |
|---|---|---|
| Provider | Alibaba (Qwen) | Alibaba (Qwen) |
| Noometry Index | 32.9 | 38.8 |
| Rank | 239 | 166 |
| Context | 33K | 131K |
| Input $/M | $0.66 | $0.70 |
| Output $/M | $1 | $2.80 |
| Weights | Open | Open |

## Coding

- Qwen2.5-Coder-32B: 22.4 (#328)
- Qwen3 32B: 37.3 (#184)

| Benchmark | Qwen2.5-Coder-32B | Qwen3 32B |
|---|---|---|
| Aider Polyglot | 16.4% | 40% |
| LMArena Coding | 1276 | 1358 |
| SWE-bench Verified (bash only) | 9% | — |
| SciCode | — | 35.4% |
| BigCodeBench Instruct | 49% | — |
| LiveBench Coding | 56.9% | — |
| BigCodeBench Complete | 58% | — |
| HumanEval+ | 87.2% | — |
| MBPP+ | 77% | — |

## Agentic & Tool Use

- Qwen2.5-Coder-32B: —
- Qwen3 32B: 32.3 (#58)

| Benchmark | Qwen2.5-Coder-32B | Qwen3 32B |
|---|---|---|
| Berkeley Function Calling Leaderboard | — | 48.7% |

## Reasoning

- Qwen2.5-Coder-32B: 20.4 (#220)
- Qwen3 32B: 19.6 (#233)

| Benchmark | Qwen2.5-Coder-32B | Qwen3 32B |
|---|---|---|
| LMArena Hard Prompts | 1251 | 1335 |
| Epoch Capabilities Index | 119.41 | 138.51 |
| Kagi LLM Benchmark | — | 54.9% |
| CritPt | — | 0.3% |
| Chess Puzzles | — | 5% |
| LiveBench Reasoning | 42.1% | — |
| DTBench | — | 67.5% |
| LiveBench Data Analysis | 49.9% | — |
| LMCA | — | 17.3% |
| HellaSwag | 83% | — |
| LiveBench | 46.2% | — |
| WinoGrande | 80.8% | — |

## Math

- Qwen2.5-Coder-32B: 30.8 (#203)
- Qwen3 32B: 37.9 (#89)

| Benchmark | Qwen2.5-Coder-32B | Qwen3 32B |
|---|---|---|
| LMArena Math | 1251 | 1399 |
| OTIS Mock AIME 2024-2025 | — | 66.9% |
| LiveBench Math | 46.6% | — |
| GSM8K | 93% | — |

## Knowledge

- Qwen2.5-Coder-32B: 32.9 (#200)
- Qwen3 32B: 40.0 (#112)

| Benchmark | Qwen2.5-Coder-32B | Qwen3 32B |
|---|---|---|
| LMArena Expert | 1221 | 1362 |
| GPQA Diamond | — | 65.7% |
| Vectara Hallucination Rate | — | 5.9% |
| ARC (AI2) Challenge | 70.5% | — |
| MMLU | 79.1% | — |

## Multilingual

- Qwen2.5-Coder-32B: 38.0 (#230)
- Qwen3 32B: 45.8 (#162)

| Benchmark | Qwen2.5-Coder-32B | Qwen3 32B |
|---|---|---|
| LMArena Non-English | 1205 | 1317 |
| LMArena Chinese | 1222 | 1357 |
| LMArena Russian | 1228 | 1311 |
| LMArena German | — | 1341 |

## Instruction Following

- Qwen2.5-Coder-32B: 61.6 (#240)
- Qwen3 32B: 69.1 (#174)

| Benchmark | Qwen2.5-Coder-32B | Qwen3 32B |
|---|---|---|
| LMArena Instruction Following | 1223 | 1305 |
| LiveBench Instruction Following | 58.7% | — |

## Long Context

- Qwen2.5-Coder-32B: 37.9 (#204)
- Qwen3 32B: 43.7 (#83)

| Benchmark | Qwen2.5-Coder-32B | Qwen3 32B |
|---|---|---|
| LMArena Longer Query | 1251 | 1327 |
| Fiction.LiveBench | — | 74.2% |

## Writing & Preference

- Qwen2.5-Coder-32B: 41.8 (#235)
- Qwen3 32B: 53.1 (#158)

| Benchmark | Qwen2.5-Coder-32B | Qwen3 32B |
|---|---|---|
| LMArena Text | 1230 | 1340 |
| LMArena Creative Writing | 1174 | 1297 |
| LMArena Multi-Turn | 1222 | 1331 |
| LiveBench Language | 23.3% | — |

## FAQ

### Is Qwen2.5-Coder-32B better than Qwen3 32B?

Qwen3 32B is the stronger model overall, scoring 38.8 to 32.9 on the Noometry Index. Qwen2.5-Coder-32B costs 1.6× less per token, which makes it the better buy when Qwen3 32B's lead doesn't matter for your workload.

### Which is cheaper, Qwen2.5-Coder-32B or Qwen3 32B?

Qwen2.5-Coder-32B is cheaper. It lists at $0.66 per million input tokens and $1 per million output tokens; Qwen3 32B lists at $0.70 and $2.80.

### Is Qwen2.5-Coder-32B or Qwen3 32B better for coding?

Qwen3 32B scores higher on coding benchmarks: 37.3 versus 22.4 in the Noometry coding category.

### Which has the bigger context window?

Qwen3 32B does, with 131K tokens against 33K.

### How many benchmarks do Qwen2.5-Coder-32B and Qwen3 32B share?

14 benchmarks have published results for both models. Qwen2.5-Coder-32B has 31 scored results on Noometry and Qwen3 32B has 26.