# Qwen2-72B vs Qwen3.5-Flash

> Qwen3.5-Flash is the stronger model overall, scoring 42.5 to 30.0 on the Noometry Index.

- Canonical page: https://noometry.com/compare/qwen2-72b-vs-qwen3-5-flash
- Last updated: 2026-10-12
- Shared benchmarks: 19

## Summary

- They share 19 benchmarks with published results for both. Qwen2-72B scores higher in 0 categories and Qwen3.5-Flash in 8 categories; 8 gaps are clear of the uncertainty.
- The widest gap is in knowledge, where Qwen3.5-Flash leads 43.2 to 21.2.
- The biggest single-benchmark swing is GPQA Diamond: 40.8% for Qwen2-72B and 82.3% for Qwen3.5-Flash.
- Qwen2-72B has downloadable open weights; the other is API-only.

## Snapshot

| | Qwen2-72B | Qwen3.5-Flash |
|---|---|---|
| Provider | Alibaba (Qwen) | Alibaba (Qwen) |
| Noometry Index | 30.0 | 42.5 |
| Rank | 300 | 112 |
| Context | — | 1M |
| Input $/M | — | $0.10 |
| Output $/M | — | $0.40 |
| Weights | Open | Proprietary |

## Coding

- Qwen2-72B: 29.1 (#310)
- Qwen3.5-Flash: 34.2 (#242)

| Benchmark | Qwen2-72B | Qwen3.5-Flash |
|---|---|---|
| LMArena Coding | 1196 | 1412 |
| LMArena WebDev | — | 1244 |
| WeirdML | 11.3% | — |
| BigCodeBench Instruct | 38.5% | — |
| BigCodeBench Complete | 54% | — |
| ALE-Bench | — | 221.8 |

## Agentic & Tool Use

- Qwen2-72B: 17.0 (#146)
- Qwen3.5-Flash: —

| Benchmark | Qwen2-72B | Qwen3.5-Flash |
|---|---|---|
| TheAgentCompany | 1.1% | — |
| METR Time Horizons | 29.9% | — |
| Vending-Bench 2 | — | 462.69 |

## Reasoning

- Qwen2-72B: 23.2 (#181)
- Qwen3.5-Flash: 33.7 (#72)

| Benchmark | Qwen2-72B | Qwen3.5-Flash |
|---|---|---|
| LMArena Hard Prompts | 1191 | 1403 |
| Epoch Capabilities Index | 125.28 | 143.98 |
| Chess Puzzles | — | 21% |
| Mystery Game Puzzles | — | 20% |
| DTBench | — | 82.9% |
| LMCA | — | 29.1% |

## Math

- Qwen2-72B: 30.2 (#236)
- Qwen3.5-Flash: 37.4 (#158)

| Benchmark | Qwen2-72B | Qwen3.5-Flash |
|---|---|---|
| LMArena Math | 1235 | 1407 |
| FrontierMath (Tiers 1-3) | — | 18.2% |
| OTIS Mock AIME 2024-2025 | — | 84.4% |
| MATH Level 5 | 39.1% | — |
| FrontierMath (Feb 2025 set) | — | 6.2% |
| FrontierMath Tier 4 (v1) | — | 0% |

## Knowledge

- Qwen2-72B: 21.2 (#275)
- Qwen3.5-Flash: 43.2 (#93)

| Benchmark | Qwen2-72B | Qwen3.5-Flash |
|---|---|---|
| GPQA Diamond | 40.8% | 82.3% |
| LMArena Expert | 1171 | 1407 |
| SimpleQA Verified | — | 20.3% |
| Vectara Hallucination Rate | — | 10.5% |
| MMLU | 82.4% | — |

## Multilingual

- Qwen2-72B: 35.9 (#244)
- Qwen3.5-Flash: 50.5 (#121)

| Benchmark | Qwen2-72B | Qwen3.5-Flash |
|---|---|---|
| LMArena Non-English | 1176 | 1385 |
| LMArena Chinese | 1240 | 1446 |
| LMArena French | 1170 | 1412 |
| LMArena German | 1151 | 1390 |
| LMArena Japanese | 1111 | 1368 |
| LMArena Korean | 1083 | 1344 |
| LMArena Russian | 1169 | 1379 |
| LMArena Spanish | 1169 | 1400 |

## Instruction Following

- Qwen2-72B: 61.7 (#241)
- Qwen3.5-Flash: 72.6 (#139)

| Benchmark | Qwen2-72B | Qwen3.5-Flash |
|---|---|---|
| LMArena Instruction Following | 1181 | 1374 |

## Long Context

- Qwen2-72B: 36.1 (#235)
- Qwen3.5-Flash: 42.4 (#124)

| Benchmark | Qwen2-72B | Qwen3.5-Flash |
|---|---|---|
| LMArena Longer Query | 1192 | 1392 |

## Writing & Preference

- Qwen2-72B: 40.8 (#241)
- Qwen3.5-Flash: 57.9 (#122)

| Benchmark | Qwen2-72B | Qwen3.5-Flash |
|---|---|---|
| LMArena Text | 1203 | 1397 |
| LMArena Creative Writing | 1181 | 1343 |
| LMArena Multi-Turn | 1196 | 1393 |

## FAQ

### Is Qwen2-72B better than Qwen3.5-Flash?

Qwen3.5-Flash is the stronger model overall, scoring 42.5 to 30.0 on the Noometry Index.

### Is Qwen2-72B or Qwen3.5-Flash better for coding?

Qwen3.5-Flash scores higher on coding benchmarks: 34.2 versus 29.1 in the Noometry coding category.

### How many benchmarks do Qwen2-72B and Qwen3.5-Flash share?

19 benchmarks have published results for both models. Qwen2-72B has 26 scored results on Noometry and Qwen3.5-Flash has 32.