# Qwen3.5 35B-A3B vs Step 3

> Qwen3.5 35B-A3B is the stronger model overall, scoring 41.7 to 39.9 on the Noometry Index.

- Canonical page: https://noometry.com/compare/qwen3-5-35b-a3b-vs-step-3
- Last updated: 2026-10-10
- Shared benchmarks: 15

## Summary

- They share 15 benchmarks with published results for both. Qwen3.5 35B-A3B scores higher in 6 categories and Step 3 in 2 categories; 8 gaps are clear of the uncertainty.
- The widest gap is in knowledge, where Qwen3.5 35B-A3B leads 47.7 to 36.2.

## Snapshot

| | Qwen3.5 35B-A3B | Step 3 |
|---|---|---|
| Provider | Alibaba (Qwen) | StepFun |
| Noometry Index | 41.7 | 39.9 |
| Rank | 118 | 145 |
| Context | 262K | — |
| Input $/M | $0.25 | — |
| Output $/M | $2 | — |
| Weights | Open | Open |

## Coding

- Qwen3.5 35B-A3B: 33.5 (#244)
- Step 3: 39.5 (#144)

| Benchmark | Qwen3.5 35B-A3B | Step 3 |
|---|---|---|
| LMArena Coding | 1410 | 1367 |
| LMArena WebDev | 1254 | — |
| SciCode | 29.3% | — |

## Reasoning

- Qwen3.5 35B-A3B: 23.9 (#158)
- Step 3: 27.6 (#102)

| Benchmark | Qwen3.5 35B-A3B | Step 3 |
|---|---|---|
| LMArena Hard Prompts | 1400 | 1355 |
| Kagi LLM Benchmark | — | 62.3% |
| CritPt | 0.6% | — |
| Chess Puzzles | 10% | — |
| DTBench | 80% | — |
| LMCA | 29.5% | — |
| Epoch Capabilities Index | 142.52 | — |

## Math

- Qwen3.5 35B-A3B: 38.4 (#86)
- Step 3: 34.9 (#148)

| Benchmark | Qwen3.5 35B-A3B | Step 3 |
|---|---|---|
| LMArena Math | 1404 | 1366 |
| MathArena Final-Answer Competitions | 56% | — |
| OTIS Mock AIME 2024-2025 | 70% | — |

## Knowledge

- Qwen3.5 35B-A3B: 47.7 (#73)
- Step 3: 36.2 (#161)

| Benchmark | Qwen3.5 35B-A3B | Step 3 |
|---|---|---|
| LMArena Expert | 1408 | 1333 |
| GPQA Diamond | 83.5% | — |
| Vectara Hallucination Rate | 10.5% | — |

## Multimodal

- Qwen3.5 35B-A3B: —
- Step 3: 34.9 (#81)

| Benchmark | Qwen3.5 35B-A3B | Step 3 |
|---|---|---|
| LMArena Vision | — | 1176 |

## Multilingual

- Qwen3.5 35B-A3B: 50.2 (#122)
- Step 3: 46.5 (#154)

| Benchmark | Qwen3.5 35B-A3B | Step 3 |
|---|---|---|
| LMArena Non-English | 1379 | 1327 |
| LMArena Chinese | 1458 | 1396 |
| LMArena German | 1367 | 1371 |
| LMArena Korean | 1354 | 1271 |
| LMArena Russian | 1377 | 1331 |
| LMArena Spanish | 1390 | 1372 |
| LMArena French | 1412 | — |
| LMArena Japanese | 1326 | — |

## Instruction Following

- Qwen3.5 35B-A3B: 73.0 (#123)
- Step 3: 70.6 (#159)

| Benchmark | Qwen3.5 35B-A3B | Step 3 |
|---|---|---|
| LMArena Instruction Following | 1379 | 1332 |

## Long Context

- Qwen3.5 35B-A3B: 42.4 (#122)
- Step 3: 40.3 (#152)

| Benchmark | Qwen3.5 35B-A3B | Step 3 |
|---|---|---|
| LMArena Longer Query | 1390 | 1326 |

## Writing & Preference

- Qwen3.5 35B-A3B: 58.1 (#118)
- Step 3: 54.5 (#146)

| Benchmark | Qwen3.5 35B-A3B | Step 3 |
|---|---|---|
| LMArena Text | 1395 | 1351 |
| LMArena Creative Writing | 1347 | 1322 |
| LMArena Multi-Turn | 1391 | 1340 |

## FAQ

### Is Qwen3.5 35B-A3B better than Step 3?

Qwen3.5 35B-A3B is the stronger model overall, scoring 41.7 to 39.9 on the Noometry Index.

### Is Qwen3.5 35B-A3B or Step 3 better for coding?

Step 3 scores higher on coding benchmarks: 39.5 versus 33.5 in the Noometry coding category.

### How many benchmarks do Qwen3.5 35B-A3B and Step 3 share?

15 benchmarks have published results for both models. Qwen3.5 35B-A3B has 28 scored results on Noometry and Step 3 has 17.