Model comparison
Qwen2.5 72B Instruct vs Qwen3.5-Flash
Qwen3.5-Flash is the stronger model overall, scoring 42.3 to 31.5 on the Noometry Index.
Last verified . 22 shared benchmarks.
Alibaba (Qwen)Summary
- They share 22 benchmarks with published results for both. Qwen2.5 72B Instruct scores higher in 0 categories and Qwen3.5-Flash in 8 categories; 8 gaps are clear of the uncertainty.
- The widest gap is in math, where Qwen3.5-Flash leads 36.3 to 18.1.
- The biggest single-benchmark swing is OTIS Mock AIME 2024-2025: 8.1% for Qwen2.5 72B Instruct and 84.4% for Qwen3.5-Flash.
- Qwen3.5-Flash is cheaper at $0.10 / $0.40 per million input/output tokens, against $1.40 / $5.60 for Qwen2.5 72B Instruct.
- Qwen3.5-Flash accepts more context: 1M tokens versus 131K.
- Qwen2.5 72B Instruct has downloadable open weights; the other is API-only.
Side by side
| Qwen2.5 72B Instruct | Qwen3.5-Flash | |
|---|---|---|
| Provider | Alibaba (Qwen) | Alibaba (Qwen) |
| Noometry Index | 31.5 | 42.3 |
| Released | 2024-09 | 2026-02-23 |
| Weights | Open | Proprietary |
| Context window | 131K | 1M |
| Max output | 8K | 66K |
| Input $ / M tokens | $1.40 | $0.10 |
| Output $ / M tokens | $5.60 | $0.40 |
| Results tracked | 43 | 32 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding Qwen3.5-Flash leads
Qwen2.5 72B Instruct: 32.8 (#256), Qwen3.5-Flash: 33.9 (#234)
| Benchmark | Qwen2.5 72B Instruct | Qwen3.5-Flash |
|---|---|---|
| LMArena Coding | 1293 | 1412 |
| LMArena WebDev | — | 1243 |
| WeirdML | 16% | — |
| BigCodeBench Instruct | 45.8% | — |
| BigCodeBench Complete | 55.9% | — |
| ALE-Bench | — | 221.8 |
Agentic & Tool Use Not comparable
Qwen2.5 72B Instruct: 22.0 (#124), Qwen3.5-Flash: —
| Benchmark | Qwen2.5 72B Instruct | Qwen3.5-Flash |
|---|---|---|
| TheAgentCompany | 5.7% | — |
| BALROG | 16.2% | — |
| METR Time Horizons | 35.8% | — |
| Vending-Bench 2 | — | 462.69 |
Reasoning Qwen3.5-Flash leads
Qwen2.5 72B Instruct: 21.4 (#197), Qwen3.5-Flash: 32.9 (#63)
| Benchmark | Qwen2.5 72B Instruct | Qwen3.5-Flash |
|---|---|---|
| LMArena Hard Prompts | 1271 | 1404 |
| DTBench | 62.9% | 82.9% |
| LMCA | 13.4% | 29.1% |
| Epoch Capabilities Index | 129.01 | 143.98 |
| Chess Puzzles | — | 21% |
| Mystery Game Puzzles | — | 20% |
| BIG-Bench Hard | 79.8% | — |
| ForecastBench | 57.5 | — |
| HellaSwag | 84.8% | — |
| PIQA | 82.6% | — |
| WinoGrande | 82.3% | — |
Math Qwen3.5-Flash leads
Qwen2.5 72B Instruct: 18.1 (#281), Qwen3.5-Flash: 36.3 (#117)
| Benchmark | Qwen2.5 72B Instruct | Qwen3.5-Flash |
|---|---|---|
| OTIS Mock AIME 2024-2025 | 8.1% | 84.4% |
| LMArena Math | 1283 | 1407 |
| FrontierMath (Tiers 1-3) | — | 18.2% |
| Omni-MATH | 33% | — |
| MATH Level 5 | 63.2% | — |
| FrontierMath (Feb 2025 set) | — | 6.2% |
| FrontierMath Tier 4 (v1) | — | 0% |
Knowledge Qwen3.5-Flash leads
Qwen2.5 72B Instruct: 26.9 (#247), Qwen3.5-Flash: 43.2 (#84)
| Benchmark | Qwen2.5 72B Instruct | Qwen3.5-Flash |
|---|---|---|
| GPQA Diamond | 49.1% | 82.3% |
| LMArena Expert | 1245 | 1408 |
| SimpleQA Verified | — | 20.3% |
| MMLU-Pro | 63.1% | — |
| Confabulations | 19.1% | — |
| Vectara Hallucination Rate | — | 10.5% |
| GPQA (HELM) | 42.6% | — |
| ARC (AI2) Challenge | 94.5% | — |
| MMLU | 85.3% | — |
| TriviaQA | 71.9% | — |
Multilingual Qwen3.5-Flash leads
Qwen2.5 72B Instruct: 41.2 (#208), Qwen3.5-Flash: 50.7 (#116)
| Benchmark | Qwen2.5 72B Instruct | Qwen3.5-Flash |
|---|---|---|
| LMArena Non-English | 1252 | 1385 |
| LMArena Chinese | 1272 | 1446 |
| LMArena French | 1279 | 1412 |
| LMArena German | 1234 | 1389 |
| LMArena Japanese | 1180 | 1369 |
| LMArena Korean | 1188 | 1344 |
| LMArena Russian | 1264 | 1379 |
| LMArena Spanish | 1255 | 1399 |
Instruction Following Qwen3.5-Flash leads
Qwen2.5 72B Instruct: 65.7 (#216), Qwen3.5-Flash: 72.8 (#133)
| Benchmark | Qwen2.5 72B Instruct | Qwen3.5-Flash |
|---|---|---|
| LMArena Instruction Following | 1254 | 1374 |
| IFEval | 80.6% | — |
Long Context Qwen3.5-Flash leads
Qwen2.5 72B Instruct: 38.9 (#183), Qwen3.5-Flash: 42.4 (#119)
| Benchmark | Qwen2.5 72B Instruct | Qwen3.5-Flash |
|---|---|---|
| LMArena Longer Query | 1282 | 1392 |
Writing & Preference Qwen3.5-Flash leads
Qwen2.5 72B Instruct: 46.9 (#210), Qwen3.5-Flash: 58.2 (#117)
| Benchmark | Qwen2.5 72B Instruct | Qwen3.5-Flash |
|---|---|---|
| LMArena Text | 1269 | 1397 |
| LMArena Creative Writing | 1221 | 1343 |
| LMArena Multi-Turn | 1272 | 1393 |
| WildBench | 80.2% | — |
Frequently asked questions
Is Qwen2.5 72B Instruct better than Qwen3.5-Flash?
Qwen3.5-Flash is the stronger model overall, scoring 42.3 to 31.5 on the Noometry Index.
Which is cheaper, Qwen2.5 72B Instruct or Qwen3.5-Flash?
Qwen3.5-Flash is cheaper. It lists at $0.10 per million input tokens and $0.40 per million output tokens; Qwen2.5 72B Instruct lists at $1.40 and $5.60.
Is Qwen2.5 72B Instruct or Qwen3.5-Flash better for coding?
Qwen3.5-Flash scores higher on coding benchmarks: 33.9 versus 32.8 in the Noometry coding category.
Which has the bigger context window?
Qwen3.5-Flash does, with 1M tokens against 131K.
How many benchmarks do Qwen2.5 72B Instruct and Qwen3.5-Flash share?
22 benchmarks have published results for both models. Qwen2.5 72B Instruct has 43 scored results on Noometry and Qwen3.5-Flash has 32.