Model comparison
Qwen3.5 397B-A17B vs Qwen3.6 27B
Qwen3.5 397B-A17B is the stronger model overall, scoring 46.0 to 42.2 on the Noometry Index.
Last verified . 8 shared benchmarks.
Alibaba (Qwen)Summary
- They share 8 benchmarks with published results for both. Qwen3.5 397B-A17B scores higher in 4 categories and Qwen3.6 27B in 1 category; 4 gaps are clear of the uncertainty.
- The widest gap is in writing & preference, where Qwen3.5 397B-A17B leads 62.3 to 50.3.
- The biggest single-benchmark swing is Mystery Game Puzzles: 18% for Qwen3.5 397B-A17B and 7% for Qwen3.6 27B.
- Both cost about the same: $0.60 input and $3.60 output per million tokens.
Side by side
| Qwen3.5 397B-A17B | Qwen3.6 27B | |
|---|---|---|
| Provider | Alibaba (Qwen) | Alibaba (Qwen) |
| Noometry Index | 46.0 | 42.2 |
| Released | 2026-02-01 | 2026-04-22 |
| Weights | Open | Open |
| Context window | 262K | 262K |
| Max output | 66K | 66K |
| Input $ / M tokens | $0.60 | $0.60 |
| Output $ / M tokens | $3.60 | $3.60 |
| Results tracked | 36 | 11 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding Qwen3.5 397B-A17B leads
Qwen3.5 397B-A17B: 42.0 (#114), Qwen3.6 27B: 39.1 (#163)
| Benchmark | Qwen3.5 397B-A17B | Qwen3.6 27B |
|---|---|---|
| LMArena WebDev | 1400 | — |
| SciCode | — | 37.3% |
| LMArena Coding | 1465 | — |
Agentic & Tool Use Not comparable
Qwen3.5 397B-A17B: 33.3 (#53), Qwen3.6 27B: —
| Benchmark | Qwen3.5 397B-A17B | Qwen3.6 27B |
|---|---|---|
| APEX-Agents | 24.9% | — |
| τ²-bench Airline | 81.5% | — |
| τ²-bench Banking | 9.8% | — |
| τ²-bench Retail | 84.4% | — |
| τ²-bench Telecom | 97.8% | — |
Reasoning Qwen3.5 397B-A17B leads
Qwen3.5 397B-A17B: 34.5 (#70), Qwen3.6 27B: 25.0 (#153)
| Benchmark | Qwen3.5 397B-A17B | Qwen3.6 27B |
|---|---|---|
| Chess Puzzles | 13% | 22% |
| Mystery Game Puzzles | 18% | 7% |
| DTBench | 87.5% | 78.1% |
| LMCA | 37.9% | 34.5% |
| Epoch Capabilities Index | 146.65 | 146.5 |
| Kagi LLM Benchmark | 73.7% | — |
| NYT Connections (extended) | 58.9% | — |
| CritPt | — | 0.9% |
| Thematic Generalization | 65.1% | — |
| LMArena Hard Prompts | 1448 | — |
Math Qwen3.6 27B leads
Qwen3.5 397B-A17B: 46.1 (#73), Qwen3.6 27B: 48.5 (#62)
| Benchmark | Qwen3.5 397B-A17B | Qwen3.6 27B |
|---|---|---|
| FrontierMath (Tiers 1-3) | 31.2% | 35.1% |
| OTIS Mock AIME 2024-2025 | 88.9% | 91.1% |
| LMArena Math | 1454 | — |
Knowledge Too close to call
Qwen3.5 397B-A17B: 53.3 (#58), Qwen3.6 27B: 52.4 (#63)
| Benchmark | Qwen3.5 397B-A17B | Qwen3.6 27B |
|---|---|---|
| GPQA Diamond | 86.4% | 85.9% |
| LMArena Expert | 1462 | — |
Multimodal Not comparable
Qwen3.5 397B-A17B: 40.7 (#44), Qwen3.6 27B: —
| Benchmark | Qwen3.5 397B-A17B | Qwen3.6 27B |
|---|---|---|
| LMArena Vision | 1263 | — |
Multilingual Not comparable
Qwen3.5 397B-A17B: 53.7 (#59), Qwen3.6 27B: —
| Benchmark | Qwen3.5 397B-A17B | Qwen3.6 27B |
|---|---|---|
| LMArena Non-English | 1430 | — |
| LMArena Chinese | 1500 | — |
| LMArena French | 1461 | — |
| LMArena German | 1447 | — |
| LMArena Japanese | 1426 | — |
| LMArena Korean | 1384 | — |
| LMArena Russian | 1429 | — |
| LMArena Spanish | 1441 | — |
Instruction Following Not comparable
Qwen3.5 397B-A17B: 75.0 (#77), Qwen3.6 27B: —
| Benchmark | Qwen3.5 397B-A17B | Qwen3.6 27B |
|---|---|---|
| LMArena Instruction Following | 1424 | — |
Long Context Not comparable
Qwen3.5 397B-A17B: 44.1 (#74), Qwen3.6 27B: —
| Benchmark | Qwen3.5 397B-A17B | Qwen3.6 27B |
|---|---|---|
| LMArena Longer Query | 1442 | — |
Writing & Preference Qwen3.5 397B-A17B leads
Qwen3.5 397B-A17B: 62.3 (#79), Qwen3.6 27B: 50.3 (#181)
| Benchmark | Qwen3.5 397B-A17B | Qwen3.6 27B |
|---|---|---|
| LMArena Text | 1438 | — |
| LMArena Creative Writing | 1401 | — |
| EQ-Bench Creative Writing | 1478 | — |
| EQ-Bench 4 | — | 1026 |
| LMArena Multi-Turn | 1446 | — |
Frequently asked questions
Is Qwen3.5 397B-A17B better than Qwen3.6 27B?
Qwen3.5 397B-A17B is the stronger model overall, scoring 46.0 to 42.2 on the Noometry Index.
Which is cheaper, Qwen3.5 397B-A17B or Qwen3.6 27B?
Qwen3.6 27B is cheaper. It lists at $0.60 per million input tokens and $3.60 per million output tokens; Qwen3.5 397B-A17B lists at $0.60 and $3.60.
Is Qwen3.5 397B-A17B or Qwen3.6 27B better for coding?
Qwen3.5 397B-A17B scores higher on coding benchmarks: 42.0 versus 39.1 in the Noometry coding category.
Which has the bigger context window?
Both accept 262K tokens.
How many benchmarks do Qwen3.5 397B-A17B and Qwen3.6 27B share?
8 benchmarks have published results for both models. Qwen3.5 397B-A17B has 36 scored results on Noometry and Qwen3.6 27B has 11.