# Qwen3.5 397B-A17B vs Step 3.5 Flash

> Qwen3.5 397B-A17B is the stronger model overall, scoring 46.0 to 42.3 on the Noometry Index. Step 3.5 Flash costs 9.0× less per token, which makes it the better buy when Qwen3.5 397B-A17B's lead doesn't matter for your workload.

- Canonical page: https://noometry.com/compare/qwen3-5-397b-a17b-vs-step-3-5-flash
- Last updated: 2026-10-10
- Shared benchmarks: 18

## Summary

- They share 18 benchmarks with published results for both. Qwen3.5 397B-A17B scores higher in 7 categories and Step 3.5 Flash in 1 category; 7 gaps are clear of the uncertainty.
- The widest gap is in knowledge, where Qwen3.5 397B-A17B leads 53.3 to 39.6.
- The biggest single-benchmark swing is NYT Connections (extended): 58.9% for Qwen3.5 397B-A17B and 28.4% for Step 3.5 Flash.
- Step 3.5 Flash is cheaper at $0.10 / $0.30 per million input/output tokens, against $0.60 / $3.60 for Qwen3.5 397B-A17B.
- Qwen3.5 397B-A17B accepts more context: 262K tokens versus 256K.

## Snapshot

| | Qwen3.5 397B-A17B | Step 3.5 Flash |
|---|---|---|
| Provider | Alibaba (Qwen) | StepFun |
| Noometry Index | 46.0 | 42.3 |
| Rank | 67 | 116 |
| Context | 262K | 256K |
| Input $/M | $0.60 | $0.10 |
| Output $/M | $3.60 | $0.30 |
| Weights | Open | Open |

## Coding

- Qwen3.5 397B-A17B: 42.0 (#114)
- Step 3.5 Flash: 42.4 (#105)

| Benchmark | Qwen3.5 397B-A17B | Step 3.5 Flash |
|---|---|---|
| LMArena Coding | 1465 | 1436 |
| LMArena WebDev | 1400 | — |

## Agentic & Tool Use

- Qwen3.5 397B-A17B: 33.3 (#53)
- Step 3.5 Flash: —

| Benchmark | Qwen3.5 397B-A17B | Step 3.5 Flash |
|---|---|---|
| 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: 34.5 (#70)
- Step 3.5 Flash: 22.2 (#202)

| Benchmark | Qwen3.5 397B-A17B | Step 3.5 Flash |
|---|---|---|
| NYT Connections (extended) | 58.9% | 28.4% |
| LMArena Hard Prompts | 1448 | 1411 |
| Kagi LLM Benchmark | 73.7% | — |
| Chess Puzzles | 13% | — |
| Thematic Generalization | 65.1% | — |
| Mystery Game Puzzles | 18% | — |
| DTBench | 87.5% | — |
| LMCA | 37.9% | — |
| Epoch Capabilities Index | 146.65 | — |

## Math

- Qwen3.5 397B-A17B: 46.1 (#73)
- Step 3.5 Flash: 42.6 (#84)

| Benchmark | Qwen3.5 397B-A17B | Step 3.5 Flash |
|---|---|---|
| LMArena Math | 1454 | 1408 |
| FrontierMath (Tiers 1-3) | 31.2% | — |
| MathArena Final-Answer Competitions | — | 66.8% |
| OTIS Mock AIME 2024-2025 | 88.9% | — |

## Knowledge

- Qwen3.5 397B-A17B: 53.3 (#58)
- Step 3.5 Flash: 39.6 (#132)

| Benchmark | Qwen3.5 397B-A17B | Step 3.5 Flash |
|---|---|---|
| LMArena Expert | 1462 | 1421 |
| GPQA Diamond | 86.4% | — |

## Multimodal

- Qwen3.5 397B-A17B: 40.7 (#44)
- Step 3.5 Flash: —

| Benchmark | Qwen3.5 397B-A17B | Step 3.5 Flash |
|---|---|---|
| LMArena Vision | 1263 | — |

## Multilingual

- Qwen3.5 397B-A17B: 53.7 (#59)
- Step 3.5 Flash: 50.5 (#119)

| Benchmark | Qwen3.5 397B-A17B | Step 3.5 Flash |
|---|---|---|
| LMArena Non-English | 1430 | 1385 |
| LMArena Chinese | 1500 | 1447 |
| LMArena French | 1461 | 1421 |
| LMArena German | 1447 | 1405 |
| LMArena Japanese | 1426 | 1354 |
| LMArena Korean | 1384 | 1352 |
| LMArena Russian | 1429 | 1385 |
| LMArena Spanish | 1441 | 1419 |

## Instruction Following

- Qwen3.5 397B-A17B: 75.0 (#77)
- Step 3.5 Flash: 73.1 (#124)

| Benchmark | Qwen3.5 397B-A17B | Step 3.5 Flash |
|---|---|---|
| LMArena Instruction Following | 1424 | 1385 |

## Long Context

- Qwen3.5 397B-A17B: 44.1 (#74)
- Step 3.5 Flash: 42.8 (#117)

| Benchmark | Qwen3.5 397B-A17B | Step 3.5 Flash |
|---|---|---|
| LMArena Longer Query | 1442 | 1402 |

## Writing & Preference

- Qwen3.5 397B-A17B: 62.3 (#79)
- Step 3.5 Flash: 58.8 (#113)

| Benchmark | Qwen3.5 397B-A17B | Step 3.5 Flash |
|---|---|---|
| LMArena Text | 1438 | 1403 |
| LMArena Creative Writing | 1401 | 1357 |
| LMArena Multi-Turn | 1446 | 1405 |
| EQ-Bench Creative Writing | 1478 | — |

## FAQ

### Is Qwen3.5 397B-A17B better than Step 3.5 Flash?

Qwen3.5 397B-A17B is the stronger model overall, scoring 46.0 to 42.3 on the Noometry Index. Step 3.5 Flash costs 9.0× less per token, which makes it the better buy when Qwen3.5 397B-A17B's lead doesn't matter for your workload.

### Which is cheaper, Qwen3.5 397B-A17B or Step 3.5 Flash?

Step 3.5 Flash is cheaper. It lists at $0.10 per million input tokens and $0.30 per million output tokens; Qwen3.5 397B-A17B lists at $0.60 and $3.60.

### Is Qwen3.5 397B-A17B or Step 3.5 Flash better for coding?

They score almost the same on coding (42.0 vs 42.4); test both on your own repository before choosing.

### Which has the bigger context window?

Qwen3.5 397B-A17B does, with 262K tokens against 256K.

### How many benchmarks do Qwen3.5 397B-A17B and Step 3.5 Flash share?

18 benchmarks have published results for both models. Qwen3.5 397B-A17B has 36 scored results on Noometry and Step 3.5 Flash has 19.