# Qwen3.6 27B vs Step 3

> Qwen3.6 27B is the stronger model overall, scoring 42.2 to 40.5 on the Noometry Index.

- Canonical page: https://noometry.com/compare/qwen3-6-27b-vs-step-3
- Last updated: 2026-10-12
- Shared benchmarks: 0

## Summary

- The widest gap is in knowledge, where Qwen3.6 27B leads 52.4 to 36.8.

## Snapshot

| | Qwen3.6 27B | Step 3 |
|---|---|---|
| Provider | Alibaba (Qwen) | StepFun |
| Noometry Index | 42.2 | 40.5 |
| Rank | 117 | 149 |
| Context | 262K | — |
| Input $/M | $0.60 | — |
| Output $/M | $3.60 | — |
| Weights | Open | Open |

## Coding

- Qwen3.6 27B: 39.1 (#163)
- Step 3: 40.1 (#147)

| Benchmark | Qwen3.6 27B | Step 3 |
|---|---|---|
| SciCode | 37.3% | — |
| LMArena Coding | — | 1367 |

## Reasoning

- Qwen3.6 27B: 25.0 (#153)
- Step 3: 28.4 (#105)

| Benchmark | Qwen3.6 27B | Step 3 |
|---|---|---|
| Kagi LLM Benchmark | — | 62.3% |
| CritPt | 0.9% | — |
| Chess Puzzles | 22% | — |
| LMArena Hard Prompts | — | 1355 |
| Mystery Game Puzzles | 7% | — |
| DTBench | 78.1% | — |
| LMCA | 34.5% | — |
| Epoch Capabilities Index | 146.5 | — |

## Math

- Qwen3.6 27B: 48.5 (#62)
- Step 3: 37.6 (#148)

| Benchmark | Qwen3.6 27B | Step 3 |
|---|---|---|
| FrontierMath (Tiers 1-3) | 35.1% | — |
| OTIS Mock AIME 2024-2025 | 91.1% | — |
| LMArena Math | — | 1366 |

## Knowledge

- Qwen3.6 27B: 52.4 (#63)
- Step 3: 36.8 (#164)

| Benchmark | Qwen3.6 27B | Step 3 |
|---|---|---|
| GPQA Diamond | 85.9% | — |
| LMArena Expert | — | 1333 |

## Multimodal

- Qwen3.6 27B: —
- Step 3: 35.5 (#86)

| Benchmark | Qwen3.6 27B | Step 3 |
|---|---|---|
| LMArena Vision | — | 1177 |

## Multilingual

- Qwen3.6 27B: —
- Step 3: 46.3 (#159)

| Benchmark | Qwen3.6 27B | Step 3 |
|---|---|---|
| LMArena Non-English | — | 1327 |
| LMArena Chinese | — | 1397 |
| LMArena German | — | 1371 |
| LMArena Korean | — | 1269 |
| LMArena Russian | — | 1331 |
| LMArena Spanish | — | 1371 |

## Instruction Following

- Qwen3.6 27B: —
- Step 3: 70.4 (#164)

| Benchmark | Qwen3.6 27B | Step 3 |
|---|---|---|
| LMArena Instruction Following | — | 1332 |

## Long Context

- Qwen3.6 27B: —
- Step 3: 40.3 (#157)

| Benchmark | Qwen3.6 27B | Step 3 |
|---|---|---|
| LMArena Longer Query | — | 1326 |

## Writing & Preference

- Qwen3.6 27B: 50.3 (#181)
- Step 3: 54.3 (#151)

| Benchmark | Qwen3.6 27B | Step 3 |
|---|---|---|
| LMArena Text | — | 1350 |
| LMArena Creative Writing | — | 1321 |
| EQ-Bench 4 | 1026 | — |
| LMArena Multi-Turn | — | 1341 |

## FAQ

### Is Qwen3.6 27B better than Step 3?

Qwen3.6 27B is the stronger model overall, scoring 42.2 to 40.5 on the Noometry Index.

### Is Qwen3.6 27B or Step 3 better for coding?

Step 3 scores higher on coding benchmarks: 40.1 versus 39.1 in the Noometry coding category.

### How many benchmarks do Qwen3.6 27B and Step 3 share?

0 benchmarks have published results for both models. Qwen3.6 27B has 11 scored results on Noometry and Step 3 has 17.