# Qwen2-72B vs Qwen3-Coder 480B-A35B Instruct

> Qwen3-Coder 480B-A35B Instruct is the stronger model overall, scoring 37.7 to 29.5 on the Noometry Index.

- Canonical page: https://noometry.com/compare/qwen2-72b-vs-qwen3-coder-480b-a35b-instruct
- Last updated: 2026-10-10
- Shared benchmarks: 18

## Summary

- They share 18 benchmarks with published results for both. Qwen2-72B scores higher in 0 categories and Qwen3-Coder 480B-A35B Instruct in 9 categories; 9 gaps are clear of the uncertainty.
- The widest gap is in knowledge, where Qwen3-Coder 480B-A35B Instruct leads 36.4 to 21.2.
- The biggest single-benchmark swing is WeirdML: 11.3% for Qwen2-72B and 41.2% for Qwen3-Coder 480B-A35B Instruct.

## Snapshot

| | Qwen2-72B | Qwen3-Coder 480B-A35B Instruct |
|---|---|---|
| Provider | Alibaba (Qwen) | Alibaba (Qwen) |
| Noometry Index | 29.5 | 37.7 |
| Rank | 294 | 186 |
| Context | — | 262K |
| Input $/M | — | $1.50 |
| Output $/M | — | $7.50 |
| Weights | Open | Open |

## Coding

- Qwen2-72B: 28.7 (#303)
- Qwen3-Coder 480B-A35B Instruct: 36.2 (#199)

| Benchmark | Qwen2-72B | Qwen3-Coder 480B-A35B Instruct |
|---|---|---|
| WeirdML | 11.3% | 41.2% |
| LMArena Coding | 1196 | 1413 |
| SWE-bench Verified (bash only) | — | 55.4% |
| LMArena WebDev | — | 1275 |
| GSO | — | 4.9% |
| BigCodeBench Instruct | 38.5% | — |
| BigCodeBench Complete | 54% | — |
| ALE-Bench | — | 461.45 |
| AlgoTune | — | 1.44 |

## Agentic & Tool Use

- Qwen2-72B: 16.9 (#138)
- Qwen3-Coder 480B-A35B Instruct: 23.7 (#116)

| Benchmark | Qwen2-72B | Qwen3-Coder 480B-A35B Instruct |
|---|---|---|
| Terminal-Bench | — | 27.2% |
| TheAgentCompany | 1.1% | — |
| METR Time Horizons | 29.9% | — |

## Reasoning

- Qwen2-72B: 22.2 (#184)
- Qwen3-Coder 480B-A35B Instruct: 24.7 (#141)

| Benchmark | Qwen2-72B | Qwen3-Coder 480B-A35B Instruct |
|---|---|---|
| LMArena Hard Prompts | 1191 | 1372 |
| Kagi LLM Benchmark | — | 49.5% |
| Epoch Capabilities Index | 125.27 | — |

## Math

- Qwen2-72B: 27.9 (#232)
- Qwen3-Coder 480B-A35B Instruct: 34.9 (#151)

| Benchmark | Qwen2-72B | Qwen3-Coder 480B-A35B Instruct |
|---|---|---|
| LMArena Math | 1235 | 1364 |
| MATH Level 5 | 39.1% | — |

## Knowledge

- Qwen2-72B: 21.2 (#270)
- Qwen3-Coder 480B-A35B Instruct: 36.4 (#159)

| Benchmark | Qwen2-72B | Qwen3-Coder 480B-A35B Instruct |
|---|---|---|
| LMArena Expert | 1171 | 1337 |
| GPQA Diamond | 40.8% | — |
| MMLU | 82.4% | — |

## Multilingual

- Qwen2-72B: 36.0 (#239)
- Qwen3-Coder 480B-A35B Instruct: 47.8 (#143)

| Benchmark | Qwen2-72B | Qwen3-Coder 480B-A35B Instruct |
|---|---|---|
| LMArena Non-English | 1176 | 1345 |
| LMArena Chinese | 1240 | 1356 |
| LMArena French | 1169 | 1394 |
| LMArena German | 1151 | 1324 |
| LMArena Japanese | 1111 | 1309 |
| LMArena Korean | 1083 | 1307 |
| LMArena Russian | 1169 | 1367 |
| LMArena Spanish | 1168 | 1360 |

## Instruction Following

- Qwen2-72B: 62.0 (#236)
- Qwen3-Coder 480B-A35B Instruct: 71.8 (#142)

| Benchmark | Qwen2-72B | Qwen3-Coder 480B-A35B Instruct |
|---|---|---|
| LMArena Instruction Following | 1181 | 1356 |

## Long Context

- Qwen2-72B: 36.1 (#230)
- Qwen3-Coder 480B-A35B Instruct: 42.0 (#126)

| Benchmark | Qwen2-72B | Qwen3-Coder 480B-A35B Instruct |
|---|---|---|
| LMArena Longer Query | 1193 | 1379 |

## Writing & Preference

- Qwen2-72B: 41.0 (#236)
- Qwen3-Coder 480B-A35B Instruct: 55.5 (#142)

| Benchmark | Qwen2-72B | Qwen3-Coder 480B-A35B Instruct |
|---|---|---|
| LMArena Text | 1203 | 1356 |
| LMArena Creative Writing | 1181 | 1333 |
| LMArena Multi-Turn | 1195 | 1365 |

## FAQ

### Is Qwen2-72B better than Qwen3-Coder 480B-A35B Instruct?

Qwen3-Coder 480B-A35B Instruct is the stronger model overall, scoring 37.7 to 29.5 on the Noometry Index.

### Is Qwen2-72B or Qwen3-Coder 480B-A35B Instruct better for coding?

Qwen3-Coder 480B-A35B Instruct scores higher on coding benchmarks: 36.2 versus 28.7 in the Noometry coding category.

### How many benchmarks do Qwen2-72B and Qwen3-Coder 480B-A35B Instruct share?

18 benchmarks have published results for both models. Qwen2-72B has 26 scored results on Noometry and Qwen3-Coder 480B-A35B Instruct has 25.