# GLM-5.3-Flash vs Mixtral 8x22B

> GLM-5.3-Flash is the stronger model overall, scoring 51.8 to 27.1 on the Noometry Index.

- Canonical page: https://noometry.com/compare/glm-5-3-flash-vs-mixtral-8x22b
- Last updated: 2026-10-11
- Shared benchmarks: 19

## Summary

- They share 19 benchmarks with published results for both. GLM-5.3-Flash scores higher in 9 categories and Mixtral 8x22B in 0 categories; 9 gaps are clear of the uncertainty.
- The widest gap is in knowledge, where GLM-5.3-Flash leads 58.4 to 15.1.
- The biggest single-benchmark swing is GPQA Diamond: 90.2% for GLM-5.3-Flash and 34.1% for Mixtral 8x22B.
- GLM-5.3-Flash is cheaper at $0.15 / $0.50 per million input/output tokens, against $2 / $6 for Mixtral 8x22B.
- GLM-5.3-Flash accepts more context: 1M tokens versus 64K.

## Snapshot

| | GLM-5.3-Flash | Mixtral 8x22B |
|---|---|---|
| Provider | Z.ai (Zhipu) | Mistral AI |
| Noometry Index | 51.8 | 27.1 |
| Rank | 41 | 333 |
| Context | 1M | 64K |
| Input $/M | $0.15 | $2 |
| Output $/M | $0.50 | $6 |
| Weights | Open | Open |

## Coding

- GLM-5.3-Flash: 53.1 (#31)
- Mixtral 8x22B: 24.2 (#329)

| Benchmark | GLM-5.3-Flash | Mixtral 8x22B |
|---|---|---|
| LMArena Coding | 1508 | 1166 |
| DeepSWE | 63.4% | — |
| FrontierCode | 31.8% | — |
| CursorBench | 36.8% | — |
| LMArena WebDev | 1609 | — |
| FrontierSWE | 18.1% | — |
| SciCode | 51.6% | — |
| WeirdML | — | 3.2% |
| BigCodeBench Instruct | — | 40.6% |
| BigCodeBench Complete | — | 50.2% |
| ALE-Bench | 303.55 | — |
| HumanEval+ | — | 72% |
| MBPP+ | — | 64.3% |

## Agentic & Tool Use

- GLM-5.3-Flash: 34.2 (#47)
- Mixtral 8x22B: 23.1 (#127)

| Benchmark | GLM-5.3-Flash | Mixtral 8x22B |
|---|---|---|
| APEX-Agents | 52.8% | — |
| Cybench | — | 7.5% |
| GDP.pdf | 14% | — |

## Reasoning

- GLM-5.3-Flash: 48.0 (#42)
- Mixtral 8x22B: 19.9 (#248)

| Benchmark | GLM-5.3-Flash | Mixtral 8x22B |
|---|---|---|
| LMArena Hard Prompts | 1491 | 1150 |
| Epoch Capabilities Index | 151.88 | 122.03 |
| ARC-AGI-2 | 65.8% | — |
| ARC-AGI-1 | 91% | — |
| CritPt | 15.4% | — |
| Chess Puzzles | 14% | — |
| Mystery Game Puzzles | 8% | — |
| DTBench | — | 55.1% |
| Surface Evolver Bench | 52.5% | — |
| Bench to the Future 3 | 0.15 | — |
| ForecastBench | — | 56.3 |

## Math

- GLM-5.3-Flash: 53.3 (#47)
- Mixtral 8x22B: 22.9 (#275)

| Benchmark | GLM-5.3-Flash | Mixtral 8x22B |
|---|---|---|
| LMArena Math | 1500 | 1184 |
| FrontierMath (Tiers 1-3) | 55.8% | — |
| FrontierMath Tier 4 | 17.1% | — |
| OTIS Mock AIME 2024-2025 | 93.9% | — |
| ProofBench | 21% | — |
| Omni-MATH | — | 16.3% |
| MATH Level 5 | — | 24.2% |

## Knowledge

- GLM-5.3-Flash: 58.4 (#36)
- Mixtral 8x22B: 15.1 (#293)

| Benchmark | GLM-5.3-Flash | Mixtral 8x22B |
|---|---|---|
| GPQA Diamond | 90.2% | 34.1% |
| LMArena Expert | 1513 | 1113 |
| MMLU-Pro | — | 46% |
| GPQA (HELM) | — | 33.4% |
| MMLU | — | 77.8% |

## Multimodal

- GLM-5.3-Flash: 42.8 (#27)
- Mixtral 8x22B: —

| Benchmark | GLM-5.3-Flash | Mixtral 8x22B |
|---|---|---|
| LMArena Vision | 1296 | — |

## Multilingual

- GLM-5.3-Flash: 56.0 (#25)
- Mixtral 8x22B: 32.8 (#255)

| Benchmark | GLM-5.3-Flash | Mixtral 8x22B |
|---|---|---|
| LMArena Non-English | 1462 | 1128 |
| LMArena Chinese | 1527 | 1116 |
| LMArena French | 1496 | 1166 |
| LMArena German | 1470 | 1141 |
| LMArena Japanese | 1429 | 1037 |
| LMArena Korean | 1446 | 1057 |
| LMArena Russian | 1469 | 1158 |
| LMArena Spanish | 1471 | 1151 |

## Instruction Following

- GLM-5.3-Flash: 77.5 (#20)
- Mixtral 8x22B: 57.7 (#266)

| Benchmark | GLM-5.3-Flash | Mixtral 8x22B |
|---|---|---|
| LMArena Instruction Following | 1478 | 1147 |
| IFEval | — | 72.4% |

## Long Context

- GLM-5.3-Flash: 45.4 (#39)
- Mixtral 8x22B: 34.7 (#247)

| Benchmark | GLM-5.3-Flash | Mixtral 8x22B |
|---|---|---|
| LMArena Longer Query | 1482 | 1144 |

## Writing & Preference

- GLM-5.3-Flash: 65.3 (#50)
- Mixtral 8x22B: 36.9 (#262)

| Benchmark | GLM-5.3-Flash | Mixtral 8x22B |
|---|---|---|
| LMArena Text | 1471 | 1162 |
| LMArena Creative Writing | 1442 | 1141 |
| LMArena Multi-Turn | 1467 | 1130 |
| WildBench | — | 71.1% |

## FAQ

### Is GLM-5.3-Flash better than Mixtral 8x22B?

GLM-5.3-Flash is the stronger model overall, scoring 51.8 to 27.1 on the Noometry Index.

### Which is cheaper, GLM-5.3-Flash or Mixtral 8x22B?

GLM-5.3-Flash is cheaper. It lists at $0.15 per million input tokens and $0.50 per million output tokens; Mixtral 8x22B lists at $2 and $6.

### Is GLM-5.3-Flash or Mixtral 8x22B better for coding?

GLM-5.3-Flash scores higher on coding benchmarks: 53.1 versus 24.2 in the Noometry coding category.

### Which has the bigger context window?

GLM-5.3-Flash does, with 1M tokens against 64K.

### How many benchmarks do GLM-5.3-Flash and Mixtral 8x22B share?

19 benchmarks have published results for both models. GLM-5.3-Flash has 40 scored results on Noometry and Mixtral 8x22B has 34.