# Qwen2.5 72B Instruct vs Qwen3.6 Flash

> Qwen3.6 Flash is the stronger model overall, scoring 38.8 to 31.9 on the Noometry Index.

- Canonical page: https://noometry.com/compare/qwen2-5-72b-instruct-vs-qwen3-6-flash
- Last updated: 2026-10-12
- Shared benchmarks: 5

## Summary

- They share 5 benchmarks with published results for both. Qwen2.5 72B Instruct scores higher in 0 categories and Qwen3.6 Flash in 3 categories; 3 gaps are clear of the uncertainty.
- The widest gap is in math, where Qwen3.6 Flash leads 39.0 to 19.3.
- The biggest single-benchmark swing is OTIS Mock AIME 2024-2025: 8.1% for Qwen2.5 72B Instruct and 84.4% for Qwen3.6 Flash.
- Qwen3.6 Flash is cheaper at $0.19 / $1.13 per million input/output tokens, against $1.40 / $5.60 for Qwen2.5 72B Instruct.
- Qwen3.6 Flash accepts more context: 1M tokens versus 131K.
- Qwen2.5 72B Instruct has downloadable open weights; the other is API-only.

## Snapshot

| | Qwen2.5 72B Instruct | Qwen3.6 Flash |
|---|---|---|
| Provider | Alibaba (Qwen) | Alibaba (Qwen) |
| Noometry Index | 31.9 | 38.8 |
| Rank | 267 | 182 |
| Context | 131K | 1M |
| Input $/M | $1.40 | $0.19 |
| Output $/M | $5.60 | $1.13 |
| Weights | Open | Proprietary |

## Coding

- Qwen2.5 72B Instruct: 33.2 (#260)
- Qwen3.6 Flash: —

| Benchmark | Qwen2.5 72B Instruct | Qwen3.6 Flash |
|---|---|---|
| WeirdML | 16% | — |
| BigCodeBench Instruct | 45.8% | — |
| LMArena Coding | 1292 | — |
| BigCodeBench Complete | 55.9% | — |
| ALE-Bench | — | 326.4 |

## Agentic & Tool Use

- Qwen2.5 72B Instruct: 22.1 (#133)
- Qwen3.6 Flash: —

| Benchmark | Qwen2.5 72B Instruct | Qwen3.6 Flash |
|---|---|---|
| TheAgentCompany | 5.7% | — |
| BALROG | 16.2% | — |
| METR Time Horizons | 35.8% | — |

## Reasoning

- Qwen2.5 72B Instruct: 22.3 (#199)
- Qwen3.6 Flash: 29.0 (#96)

| Benchmark | Qwen2.5 72B Instruct | Qwen3.6 Flash |
|---|---|---|
| DTBench | 62.9% | 77.1% |
| LMCA | 13.4% | 31% |
| Epoch Capabilities Index | 129 | 143.26 |
| SimpleBench | — | 35.2% |
| Chess Puzzles | — | 20% |
| LMArena Hard Prompts | 1271 | — |
| Mystery Game Puzzles | — | 18% |
| BIG-Bench Hard | 79.8% | — |
| ForecastBench | 57.5 | — |
| HellaSwag | 84.8% | — |
| PIQA | 82.6% | — |
| WinoGrande | 82.3% | — |

## Math

- Qwen2.5 72B Instruct: 19.3 (#287)
- Qwen3.6 Flash: 39.0 (#117)

| Benchmark | Qwen2.5 72B Instruct | Qwen3.6 Flash |
|---|---|---|
| OTIS Mock AIME 2024-2025 | 8.1% | 84.4% |
| FrontierMath (Tiers 1-3) | — | 22.5% |
| Omni-MATH | 33% | — |
| LMArena Math | 1283 | — |
| MATH Level 5 | 63.2% | — |
| FrontierMath (Feb 2025 set) | — | 10.3% |
| FrontierMath Tier 4 (v1) | — | 0% |

## Knowledge

- Qwen2.5 72B Instruct: 27.0 (#253)
- Qwen3.6 Flash: 42.1 (#100)

| Benchmark | Qwen2.5 72B Instruct | Qwen3.6 Flash |
|---|---|---|
| GPQA Diamond | 49.1% | 83.3% |
| SimpleQA Verified | — | 15.9% |
| MMLU-Pro | 63.1% | — |
| Confabulations | 19.1% | — |
| GPQA (HELM) | 42.6% | — |
| LMArena Expert | 1245 | — |
| ARC (AI2) Challenge | 94.5% | — |
| MMLU | 85.3% | — |
| TriviaQA | 71.9% | — |

## Multilingual

- Qwen2.5 72B Instruct: 41.0 (#213)
- Qwen3.6 Flash: —

| Benchmark | Qwen2.5 72B Instruct | Qwen3.6 Flash |
|---|---|---|
| LMArena Non-English | 1252 | — |
| LMArena Chinese | 1272 | — |
| LMArena French | 1280 | — |
| LMArena German | 1234 | — |
| LMArena Japanese | 1180 | — |
| LMArena Korean | 1188 | — |
| LMArena Russian | 1264 | — |
| LMArena Spanish | 1256 | — |

## Instruction Following

- Qwen2.5 72B Instruct: 65.5 (#221)
- Qwen3.6 Flash: —

| Benchmark | Qwen2.5 72B Instruct | Qwen3.6 Flash |
|---|---|---|
| IFEval | 80.6% | — |
| LMArena Instruction Following | 1254 | — |

## Long Context

- Qwen2.5 72B Instruct: 38.9 (#188)
- Qwen3.6 Flash: —

| Benchmark | Qwen2.5 72B Instruct | Qwen3.6 Flash |
|---|---|---|
| LMArena Longer Query | 1282 | — |

## Writing & Preference

- Qwen2.5 72B Instruct: 46.7 (#215)
- Qwen3.6 Flash: —

| Benchmark | Qwen2.5 72B Instruct | Qwen3.6 Flash |
|---|---|---|
| LMArena Text | 1269 | — |
| LMArena Creative Writing | 1221 | — |
| WildBench | 80.2% | — |
| LMArena Multi-Turn | 1272 | — |

## FAQ

### Is Qwen2.5 72B Instruct better than Qwen3.6 Flash?

Qwen3.6 Flash is the stronger model overall, scoring 38.8 to 31.9 on the Noometry Index.

### Which is cheaper, Qwen2.5 72B Instruct or Qwen3.6 Flash?

Qwen3.6 Flash is cheaper. It lists at $0.19 per million input tokens and $1.13 per million output tokens; Qwen2.5 72B Instruct lists at $1.40 and $5.60.

### Which has the bigger context window?

Qwen3.6 Flash does, with 1M tokens against 131K.

### How many benchmarks do Qwen2.5 72B Instruct and Qwen3.6 Flash share?

5 benchmarks have published results for both models. Qwen2.5 72B Instruct has 43 scored results on Noometry and Qwen3.6 Flash has 13.