Model comparison
Qwen3.5 27B vs Qwen3.5-Flash
Qwen3.5 27B and Qwen3.5-Flash score almost the same on the Noometry Index (41.9 vs 42.5), so choose on price, context window or the category you care about most.
Last verified . 23 shared benchmarks.
Alibaba (Qwen)Summary
- They share 23 benchmarks with published results for both. Qwen3.5 27B scores higher in 6 categories and Qwen3.5-Flash in 2 categories; 5 gaps are clear of the uncertainty.
- The widest gap is in reasoning, where Qwen3.5-Flash leads 33.7 to 27.5.
- Qwen3.5-Flash is cheaper at $0.10 / $0.40 per million input/output tokens, against $0.30 / $2.40 for Qwen3.5 27B.
- Qwen3.5-Flash accepts more context: 1M tokens versus 262K.
- Qwen3.5 27B has downloadable open weights; the other is API-only.
Side by side
| Qwen3.5 27B | Qwen3.5-Flash | |
|---|---|---|
| Provider | Alibaba (Qwen) | Alibaba (Qwen) |
| Noometry Index | 41.9 | 42.5 |
| Released | 2026-02-23 | 2026-02-23 |
| Weights | Open | Proprietary |
| Context window | 262K | 1M |
| Max output | 66K | 66K |
| Input $ / M tokens | $0.30 | $0.10 |
| Output $ / M tokens | $2.40 | $0.40 |
| Results tracked | 28 | 32 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding Qwen3.5 27B leads
Qwen3.5 27B: 38.9 (#168), Qwen3.5-Flash: 34.2 (#242)
| Benchmark | Qwen3.5 27B | Qwen3.5-Flash |
|---|---|---|
| LMArena WebDev | 1358 | 1244 |
| LMArena Coding | 1427 | 1412 |
| ALE-Bench | 349.45 | 221.8 |
| WeirdML | 39.5% | — |
Agentic & Tool Use Not comparable
Qwen3.5 27B: —, Qwen3.5-Flash: —
| Benchmark | Qwen3.5 27B | Qwen3.5-Flash |
|---|---|---|
| Vending-Bench 2 | 201.98 | 462.69 |
Reasoning Qwen3.5-Flash leads
Qwen3.5 27B: 27.5 (#117), Qwen3.5-Flash: 33.7 (#72)
| Benchmark | Qwen3.5 27B | Qwen3.5-Flash |
|---|---|---|
| LMArena Hard Prompts | 1414 | 1403 |
| DTBench | 82.4% | 82.9% |
| LMCA | 34% | 29.1% |
| NYT Connections (extended) | 47.9% | — |
| Chess Puzzles | — | 21% |
| Thematic Generalization | 45.5% | — |
| Mystery Game Puzzles | — | 20% |
| Epoch Capabilities Index | — | 143.98 |
Math Qwen3.5 27B leads
Qwen3.5 27B: 38.8 (#127), Qwen3.5-Flash: 37.4 (#158)
| Benchmark | Qwen3.5 27B | Qwen3.5-Flash |
|---|---|---|
| LMArena Math | 1429 | 1407 |
| FrontierMath (Tiers 1-3) | — | 18.2% |
| MathArena Final-Answer Competitions | 56.7% | — |
| OTIS Mock AIME 2024-2025 | — | 84.4% |
| FrontierMath (Feb 2025 set) | — | 6.2% |
| FrontierMath Tier 4 (v1) | — | 0% |
Knowledge Qwen3.5-Flash leads
Qwen3.5 27B: 38.0 (#150), Qwen3.5-Flash: 43.2 (#93)
| Benchmark | Qwen3.5 27B | Qwen3.5-Flash |
|---|---|---|
| Vectara Hallucination Rate | 12.1% | 10.5% |
| LMArena Expert | 1428 | 1407 |
| GPQA Diamond | — | 82.3% |
| SimpleQA Verified | — | 20.3% |
Multimodal Not comparable
Qwen3.5 27B: 39.4 (#59), Qwen3.5-Flash: —
| Benchmark | Qwen3.5 27B | Qwen3.5-Flash |
|---|---|---|
| LMArena Vision | 1241 | — |
Multilingual Too close to call
Qwen3.5 27B: 50.8 (#115), Qwen3.5-Flash: 50.5 (#121)
| Benchmark | Qwen3.5 27B | Qwen3.5-Flash |
|---|---|---|
| LMArena Non-English | 1390 | 1385 |
| LMArena Chinese | 1478 | 1446 |
| LMArena French | 1410 | 1412 |
| LMArena German | 1393 | 1390 |
| LMArena Japanese | 1345 | 1368 |
| LMArena Korean | 1358 | 1344 |
| LMArena Russian | 1390 | 1379 |
| LMArena Spanish | 1407 | 1400 |
Instruction Following Too close to call
Qwen3.5 27B: 73.5 (#119), Qwen3.5-Flash: 72.6 (#139)
| Benchmark | Qwen3.5 27B | Qwen3.5-Flash |
|---|---|---|
| LMArena Instruction Following | 1393 | 1374 |
Long Context Too close to call
Qwen3.5 27B: 43.1 (#106), Qwen3.5-Flash: 42.4 (#124)
| Benchmark | Qwen3.5 27B | Qwen3.5-Flash |
|---|---|---|
| LMArena Longer Query | 1413 | 1392 |
Writing & Preference Qwen3.5 27B leads
Qwen3.5 27B: 59.3 (#111), Qwen3.5-Flash: 57.9 (#122)
| Benchmark | Qwen3.5 27B | Qwen3.5-Flash |
|---|---|---|
| LMArena Text | 1409 | 1397 |
| LMArena Creative Writing | 1362 | 1343 |
| LMArena Multi-Turn | 1410 | 1393 |
Frequently asked questions
Is Qwen3.5 27B better than Qwen3.5-Flash?
Qwen3.5 27B and Qwen3.5-Flash score almost the same on the Noometry Index (41.9 vs 42.5), so choose on price, context window or the category you care about most.
Which is cheaper, Qwen3.5 27B 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; Qwen3.5 27B lists at $0.30 and $2.40.
Is Qwen3.5 27B or Qwen3.5-Flash better for coding?
Qwen3.5 27B scores higher on coding benchmarks: 38.9 versus 34.2 in the Noometry coding category.
Which has the bigger context window?
Qwen3.5-Flash does, with 1M tokens against 262K.
How many benchmarks do Qwen3.5 27B and Qwen3.5-Flash share?
23 benchmarks have published results for both models. Qwen3.5 27B has 28 scored results on Noometry and Qwen3.5-Flash has 32.