# GPT-6 Luna vs Mixtral 8x22B

> GPT-6 Luna is the stronger model overall, scoring 53.3 to 27.1 on the Noometry Index.

- Canonical page: https://noometry.com/compare/gpt-6-luna-vs-mixtral-8x22b
- Last updated: 2026-10-12
- Shared benchmarks: 20

## Summary

- They share 20 benchmarks with published results for both. GPT-6 Luna scores higher in 9 categories and Mixtral 8x22B in 0 categories; 9 gaps are clear of the uncertainty.
- The widest gap is in math, where GPT-6 Luna leads 76.1 to 22.9.
- The biggest single-benchmark swing is GPQA Diamond: 90.5% for GPT-6 Luna and 34.1% for Mixtral 8x22B.
- GPT-6 Luna is cheaper at $0.10 / $0.50 per million input/output tokens, against $2 / $6 for Mixtral 8x22B.
- GPT-6 Luna accepts more context: 1.05M tokens versus 64K.
- Mixtral 8x22B has downloadable open weights; the other is API-only.

## Snapshot

| | GPT-6 Luna | Mixtral 8x22B |
|---|---|---|
| Provider | OpenAI | Mistral AI |
| Noometry Index | 53.3 | 27.1 |
| Rank | 36 | 333 |
| Context | 1.05M | 64K |
| Input $/M | $0.10 | $2 |
| Output $/M | $0.50 | $6 |
| Weights | Proprietary | Open |

## Coding

- GPT-6 Luna: 55.5 (#25)
- Mixtral 8x22B: 24.2 (#329)

| Benchmark | GPT-6 Luna | Mixtral 8x22B |
|---|---|---|
| LMArena Coding | 1439 | 1166 |
| DeepSWE | 66.6% | — |
| FrontierCode | 42.4% | — |
| LMArena WebDev | 1581 | — |
| SciCode | 54.6% | — |
| WeirdML | — | 3.2% |
| BigCodeBench Instruct | — | 40.6% |
| BigCodeBench Complete | — | 50.2% |
| ALE-Bench | 1,577 | — |
| HumanEval+ | — | 72% |
| MBPP+ | — | 64.3% |

## Agentic & Tool Use

- GPT-6 Luna: 33.3 (#54)
- Mixtral 8x22B: 23.1 (#127)

| Benchmark | GPT-6 Luna | Mixtral 8x22B |
|---|---|---|
| APEX-Agents | 44.3% | — |
| Cybench | — | 7.5% |
| GDP.pdf | 23% | — |

## Reasoning

- GPT-6 Luna: 48.2 (#41)
- Mixtral 8x22B: 19.9 (#248)

| Benchmark | GPT-6 Luna | Mixtral 8x22B |
|---|---|---|
| LMArena Hard Prompts | 1411 | 1150 |
| DTBench | 90.1% | 55.1% |
| Epoch Capabilities Index | 156.28 | 122.03 |
| ARC-AGI-2 | 59.3% | — |
| NYT Connections (extended) | 68.7% | — |
| ARC-AGI-1 | 86.7% | — |
| CritPt | 19.4% | — |
| Chess Puzzles | 31% | — |
| Mystery Game Puzzles | 7% | — |
| LMCA | 44.5% | — |
| ForecastBench | — | 56.3 |

## Math

- GPT-6 Luna: 76.1 (#15)
- Mixtral 8x22B: 22.9 (#275)

| Benchmark | GPT-6 Luna | Mixtral 8x22B |
|---|---|---|
| LMArena Math | 1416 | 1184 |
| FrontierMath (Tiers 1-3) | 78.9% | — |
| FrontierMath Tier 4 | 56.1% | — |
| OTIS Mock AIME 2024-2025 | 98.9% | — |
| ProofBench | 64% | — |
| Omni-MATH | — | 16.3% |
| MATH Level 5 | — | 24.2% |

## Knowledge

- GPT-6 Luna: 57.0 (#41)
- Mixtral 8x22B: 15.1 (#293)

| Benchmark | GPT-6 Luna | Mixtral 8x22B |
|---|---|---|
| GPQA Diamond | 90.5% | 34.1% |
| LMArena Expert | 1444 | 1113 |
| SimpleQA Verified | 41.4% | — |
| MMLU-Pro | — | 46% |
| GPQA (HELM) | — | 33.4% |
| MMLU | — | 77.8% |

## Multimodal

- GPT-6 Luna: 42.4 (#30)
- Mixtral 8x22B: —

| Benchmark | GPT-6 Luna | Mixtral 8x22B |
|---|---|---|
| LMArena Vision | 1217 | — |
| Blueprint-Bench 2 | 31.2% | — |
| Furniture Assembly | 44.2% | — |

## Multilingual

- GPT-6 Luna: 50.5 (#117)
- Mixtral 8x22B: 32.8 (#255)

| Benchmark | GPT-6 Luna | Mixtral 8x22B |
|---|---|---|
| LMArena Non-English | 1386 | 1128 |
| LMArena Chinese | 1433 | 1116 |
| LMArena French | 1420 | 1166 |
| LMArena German | 1369 | 1141 |
| LMArena Japanese | 1369 | 1037 |
| LMArena Korean | 1360 | 1057 |
| LMArena Russian | 1394 | 1158 |
| LMArena Spanish | 1393 | 1151 |

## Instruction Following

- GPT-6 Luna: 74.3 (#99)
- Mixtral 8x22B: 57.7 (#266)

| Benchmark | GPT-6 Luna | Mixtral 8x22B |
|---|---|---|
| LMArena Instruction Following | 1409 | 1147 |
| IFEval | — | 72.4% |

## Long Context

- GPT-6 Luna: 43.0 (#111)
- Mixtral 8x22B: 34.7 (#247)

| Benchmark | GPT-6 Luna | Mixtral 8x22B |
|---|---|---|
| LMArena Longer Query | 1409 | 1144 |

## Writing & Preference

- GPT-6 Luna: 58.3 (#119)
- Mixtral 8x22B: 36.9 (#262)

| Benchmark | GPT-6 Luna | Mixtral 8x22B |
|---|---|---|
| LMArena Text | 1391 | 1162 |
| LMArena Creative Writing | 1363 | 1141 |
| LMArena Multi-Turn | 1396 | 1130 |
| WildBench | — | 71.1% |

## FAQ

### Is GPT-6 Luna better than Mixtral 8x22B?

GPT-6 Luna is the stronger model overall, scoring 53.3 to 27.1 on the Noometry Index.

### Which is cheaper, GPT-6 Luna or Mixtral 8x22B?

GPT-6 Luna is cheaper. It lists at $0.10 per million input tokens and $0.50 per million output tokens; Mixtral 8x22B lists at $2 and $6.

### Is GPT-6 Luna or Mixtral 8x22B better for coding?

GPT-6 Luna scores higher on coding benchmarks: 55.5 versus 24.2 in the Noometry coding category.

### Which has the bigger context window?

GPT-6 Luna does, with 1.05M tokens against 64K.

### How many benchmarks do GPT-6 Luna and Mixtral 8x22B share?

20 benchmarks have published results for both models. GPT-6 Luna has 42 scored results on Noometry and Mixtral 8x22B has 34.