# GLM-4.6V vs Mixtral 8x7B

> GLM-4.6V is the stronger model overall, scoring 40.9 to 26.8 on the Noometry Index.

- Canonical page: https://noometry.com/compare/glm-4-6v-vs-mixtral-8x7b
- Last updated: 2026-10-10
- Shared benchmarks: 11

## Summary

- They share 11 benchmarks with published results for both. GLM-4.6V scores higher in 7 categories and Mixtral 8x7B in 0 categories; 7 gaps are clear of the uncertainty.
- The widest gap is in knowledge, where GLM-4.6V leads 37.4 to 11.0.
- GLM-4.6V is cheaper at $0.30 / $0.90 per million input/output tokens, against $0.70 / $0.70 for Mixtral 8x7B.
- GLM-4.6V accepts more context: 128K tokens versus 32K.

## Snapshot

| | GLM-4.6V | Mixtral 8x7B |
|---|---|---|
| Provider | Z.ai (Zhipu) | Mistral AI |
| Noometry Index | 40.9 | 26.8 |
| Rank | 133 | 328 |
| Context | 128K | 32K |
| Input $/M | $0.30 | $0.70 |
| Output $/M | $0.90 | $0.70 |
| Weights | Open | Open |

## Coding

- GLM-4.6V: 40.2 (#128)
- Mixtral 8x7B: 32.2 (#265)

| Benchmark | GLM-4.6V | Mixtral 8x7B |
|---|---|---|
| LMArena Coding | 1390 | 1127 |
| HumanEval+ | — | 39.6% |
| MBPP+ | — | 49.7% |

## Reasoning

- GLM-4.6V: 26.5 (#117)
- Mixtral 8x7B: 17.4 (#283)

| Benchmark | GLM-4.6V | Mixtral 8x7B |
|---|---|---|
| LMArena Hard Prompts | 1368 | 1115 |
| DTBench | — | 49.6% |
| Adversarial NLI | — | 55.2% |
| Epoch Capabilities Index | — | 118.38 |
| ForecastBench | — | 56.3 |
| HellaSwag | — | 86.7% |
| PIQA | — | 83.6% |
| WinoGrande | — | 77.2% |

## Math

- GLM-4.6V: —
- Mixtral 8x7B: 17.2 (#284)

| Benchmark | GLM-4.6V | Mixtral 8x7B |
|---|---|---|
| Omni-MATH | — | 10.5% |
| LMArena Math | — | 1147 |
| MATH Level 5 | — | 10% |
| GSM8K | — | 74.4% |

## Knowledge

- GLM-4.6V: 37.4 (#148)
- Mixtral 8x7B: 11.0 (#296)

| Benchmark | GLM-4.6V | Mixtral 8x7B |
|---|---|---|
| LMArena Expert | 1372 | 1088 |
| GPQA Diamond | — | 30.6% |
| MMLU-Pro | — | 33.5% |
| GPQA (HELM) | — | 29.6% |
| ARC (AI2) Challenge | — | 87.3% |
| MMLU | — | 70.6% |
| OpenBookQA | — | 85.8% |
| TriviaQA | — | 82.2% |

## Multimodal

- GLM-4.6V: 34.2 (#83)
- Mixtral 8x7B: —

| Benchmark | GLM-4.6V | Mixtral 8x7B |
|---|---|---|
| LMArena Vision | 1164 | — |

## Multilingual

- GLM-4.6V: 48.8 (#135)
- Mixtral 8x7B: 29.8 (#261)

| Benchmark | GLM-4.6V | Mixtral 8x7B |
|---|---|---|
| LMArena Non-English | 1359 | 1077 |
| LMArena Chinese | 1426 | 1055 |
| LMArena Russian | 1341 | 1090 |
| LMArena French | — | 1165 |
| LMArena German | — | 1114 |
| LMArena Japanese | — | 931 |
| LMArena Korean | — | 968 |
| LMArena Spanish | — | 1110 |

## Instruction Following

- GLM-4.6V: 71.6 (#145)
- Mixtral 8x7B: 51.2 (#292)

| Benchmark | GLM-4.6V | Mixtral 8x7B |
|---|---|---|
| LMArena Instruction Following | 1351 | 1110 |
| IFEval | — | 57.5% |

## Long Context

- GLM-4.6V: 41.3 (#138)
- Mixtral 8x7B: 33.4 (#255)

| Benchmark | GLM-4.6V | Mixtral 8x7B |
|---|---|---|
| LMArena Longer Query | 1358 | 1103 |

## Writing & Preference

- GLM-4.6V: 56.8 (#132)
- Mixtral 8x7B: 34.4 (#265)

| Benchmark | GLM-4.6V | Mixtral 8x7B |
|---|---|---|
| LMArena Text | 1377 | 1132 |
| LMArena Creative Writing | 1347 | 1109 |
| LMArena Multi-Turn | 1359 | 1115 |
| WildBench | — | 67.3% |

## FAQ

### Is GLM-4.6V better than Mixtral 8x7B?

GLM-4.6V is the stronger model overall, scoring 40.9 to 26.8 on the Noometry Index.

### Which is cheaper, GLM-4.6V or Mixtral 8x7B?

GLM-4.6V is cheaper. It lists at $0.30 per million input tokens and $0.90 per million output tokens; Mixtral 8x7B lists at $0.70 and $0.70.

### Is GLM-4.6V or Mixtral 8x7B better for coding?

GLM-4.6V scores higher on coding benchmarks: 40.2 versus 32.2 in the Noometry coding category.

### Which has the bigger context window?

GLM-4.6V does, with 128K tokens against 32K.

### How many benchmarks do GLM-4.6V and Mixtral 8x7B share?

11 benchmarks have published results for both models. GLM-4.6V has 12 scored results on Noometry and Mixtral 8x7B has 38.