# Mixtral 8x7B vs Qwen3.5 27B

> Qwen3.5 27B is the stronger model overall, scoring 41.9 to 27.1 on the Noometry Index.

- Canonical page: https://noometry.com/compare/mixtral-8x7b-vs-qwen3-5-27b
- Last updated: 2026-10-12
- Shared benchmarks: 18

## Summary

- They share 18 benchmarks with published results for both. Mixtral 8x7B scores higher in 0 categories and Qwen3.5 27B in 8 categories; 8 gaps are clear of the uncertainty.
- The widest gap is in knowledge, where Qwen3.5 27B leads 38.0 to 11.0.
- The biggest single-benchmark swing is DTBench: 49.6% for Mixtral 8x7B and 82.4% for Qwen3.5 27B.
- Mixtral 8x7B is cheaper at $0.70 / $0.70 per million input/output tokens, against $0.30 / $2.40 for Qwen3.5 27B.
- Qwen3.5 27B accepts more context: 262K tokens versus 32K.

## Snapshot

| | Mixtral 8x7B | Qwen3.5 27B |
|---|---|---|
| Provider | Mistral AI | Alibaba (Qwen) |
| Noometry Index | 27.1 | 41.9 |
| Rank | 334 | 127 |
| Context | 32K | 262K |
| Input $/M | $0.70 | $0.30 |
| Output $/M | $0.70 | $2.40 |
| Weights | Open | Open |

## Coding

- Mixtral 8x7B: 32.8 (#269)
- Qwen3.5 27B: 38.9 (#168)

| Benchmark | Mixtral 8x7B | Qwen3.5 27B |
|---|---|---|
| LMArena Coding | 1126 | 1427 |
| LMArena WebDev | — | 1358 |
| WeirdML | — | 39.5% |
| ALE-Bench | — | 349.45 |
| HumanEval+ | 39.6% | — |
| MBPP+ | 49.7% | — |

## Agentic & Tool Use

- Mixtral 8x7B: —
- Qwen3.5 27B: —

| Benchmark | Mixtral 8x7B | Qwen3.5 27B |
|---|---|---|
| Vending-Bench 2 | — | 201.98 |

## Reasoning

- Mixtral 8x7B: 18.2 (#285)
- Qwen3.5 27B: 27.5 (#117)

| Benchmark | Mixtral 8x7B | Qwen3.5 27B |
|---|---|---|
| LMArena Hard Prompts | 1115 | 1414 |
| DTBench | 49.6% | 82.4% |
| NYT Connections (extended) | — | 47.9% |
| Thematic Generalization | — | 45.5% |
| LMCA | — | 34% |
| Adversarial NLI | 55.2% | — |
| Epoch Capabilities Index | 118.47 | — |
| ForecastBench | 56.3 | — |
| HellaSwag | 86.7% | — |
| PIQA | 83.6% | — |
| WinoGrande | 77.2% | — |

## Math

- Mixtral 8x7B: 18.8 (#289)
- Qwen3.5 27B: 38.8 (#127)

| Benchmark | Mixtral 8x7B | Qwen3.5 27B |
|---|---|---|
| LMArena Math | 1147 | 1429 |
| MathArena Final-Answer Competitions | — | 56.7% |
| Omni-MATH | 10.5% | — |
| MATH Level 5 | 10% | — |
| GSM8K | 74.4% | — |

## Knowledge

- Mixtral 8x7B: 11.0 (#301)
- Qwen3.5 27B: 38.0 (#150)

| Benchmark | Mixtral 8x7B | Qwen3.5 27B |
|---|---|---|
| LMArena Expert | 1088 | 1428 |
| GPQA Diamond | 30.6% | — |
| MMLU-Pro | 33.5% | — |
| Vectara Hallucination Rate | — | 12.1% |
| GPQA (HELM) | 29.6% | — |
| ARC (AI2) Challenge | 87.3% | — |
| MMLU | 70.6% | — |
| OpenBookQA | 85.8% | — |
| TriviaQA | 82.2% | — |

## Multimodal

- Mixtral 8x7B: —
- Qwen3.5 27B: 39.4 (#59)

| Benchmark | Mixtral 8x7B | Qwen3.5 27B |
|---|---|---|
| LMArena Vision | — | 1241 |

## Multilingual

- Mixtral 8x7B: 29.6 (#266)
- Qwen3.5 27B: 50.8 (#115)

| Benchmark | Mixtral 8x7B | Qwen3.5 27B |
|---|---|---|
| LMArena Non-English | 1077 | 1390 |
| LMArena Chinese | 1055 | 1478 |
| LMArena French | 1166 | 1410 |
| LMArena German | 1114 | 1393 |
| LMArena Japanese | 931 | 1345 |
| LMArena Korean | 968 | 1358 |
| LMArena Russian | 1090 | 1390 |
| LMArena Spanish | 1111 | 1407 |

## Instruction Following

- Mixtral 8x7B: 51.0 (#297)
- Qwen3.5 27B: 73.5 (#119)

| Benchmark | Mixtral 8x7B | Qwen3.5 27B |
|---|---|---|
| LMArena Instruction Following | 1109 | 1393 |
| IFEval | 57.5% | — |

## Long Context

- Mixtral 8x7B: 33.4 (#260)
- Qwen3.5 27B: 43.1 (#106)

| Benchmark | Mixtral 8x7B | Qwen3.5 27B |
|---|---|---|
| LMArena Longer Query | 1103 | 1413 |

## Writing & Preference

- Mixtral 8x7B: 34.2 (#270)
- Qwen3.5 27B: 59.3 (#111)

| Benchmark | Mixtral 8x7B | Qwen3.5 27B |
|---|---|---|
| LMArena Text | 1132 | 1409 |
| LMArena Creative Writing | 1109 | 1362 |
| LMArena Multi-Turn | 1115 | 1410 |
| WildBench | 67.3% | — |

## FAQ

### Is Mixtral 8x7B better than Qwen3.5 27B?

Qwen3.5 27B is the stronger model overall, scoring 41.9 to 27.1 on the Noometry Index.

### Which is cheaper, Mixtral 8x7B or Qwen3.5 27B?

Mixtral 8x7B is cheaper. It lists at $0.70 per million input tokens and $0.70 per million output tokens; Qwen3.5 27B lists at $0.30 and $2.40.

### Is Mixtral 8x7B or Qwen3.5 27B better for coding?

Qwen3.5 27B scores higher on coding benchmarks: 38.9 versus 32.8 in the Noometry coding category.

### Which has the bigger context window?

Qwen3.5 27B does, with 262K tokens against 32K.

### How many benchmarks do Mixtral 8x7B and Qwen3.5 27B share?

18 benchmarks have published results for both models. Mixtral 8x7B has 38 scored results on Noometry and Qwen3.5 27B has 28.