# Codestral vs GPT-6 Luna

> GPT-6 Luna is the stronger model overall, scoring 45.1 to 30.1 on the Noometry Index.

- Canonical page: https://noometry.com/compare/codestral-vs-gpt-6-luna
- Last updated: 2026-10-10
- Shared benchmarks: 0

## Summary

- The widest gap is in coding, where GPT-6 Luna leads 54.7 to 26.9.
- GPT-6 Luna is cheaper at $0.10 / $0.50 per million input/output tokens, against $0.30 / $0.90 for Codestral.
- GPT-6 Luna accepts more context: 1.05M tokens versus 256K.
- Codestral has downloadable open weights; the other is API-only.

## Snapshot

| | Codestral | GPT-6 Luna |
|---|---|---|
| Provider | Mistral AI | OpenAI |
| Noometry Index | 30.1 | 45.1 |
| Rank | 287 | 65 |
| Context | 256K | 1.05M |
| Input $/M | $0.30 | $0.10 |
| Output $/M | $0.90 | $0.50 |
| Weights | Open | Proprietary |

## Coding

- Codestral: 26.9 (#316)
- GPT-6 Luna: 54.7 (#24)

| Benchmark | Codestral | GPT-6 Luna |
|---|---|---|
| DeepSWE | — | 66.6% |
| FrontierCode | — | 42.4% |
| Aider Polyglot | 11.1% | — |
| LMArena WebDev | — | 1583 |
| BigCodeBench Instruct | 41.8% | — |
| LMArena Coding | — | 1448 |
| BigCodeBench Complete | 52.5% | — |
| ALE-Bench | 137.78 | — |
| HumanEval+ | 73.8% | — |
| MBPP+ | 61.9% | — |

## Reasoning

- Codestral: 19.1 (#244)
- GPT-6 Luna: 31.7 (#68)

| Benchmark | Codestral | GPT-6 Luna |
|---|---|---|
| Kagi LLM Benchmark | 32.5% | — |
| NYT Connections (extended) | — | 68.7% |
| LMArena Hard Prompts | — | 1415 |

## Math

- Codestral: —
- GPT-6 Luna: 35.6 (#140)

| Benchmark | Codestral | GPT-6 Luna |
|---|---|---|
| LMArena Math | — | 1396 |

## Knowledge

- Codestral: —
- GPT-6 Luna: 40.1 (#110)

| Benchmark | Codestral | GPT-6 Luna |
|---|---|---|
| LMArena Expert | — | 1458 |

## Multimodal

- Codestral: —
- GPT-6 Luna: 42.0 (#27)

| Benchmark | Codestral | GPT-6 Luna |
|---|---|---|
| Furniture Assembly | — | 44.2% |

## Multilingual

- Codestral: —
- GPT-6 Luna: 51.0 (#111)

| Benchmark | Codestral | GPT-6 Luna |
|---|---|---|
| LMArena Non-English | — | 1390 |
| LMArena Chinese | — | 1425 |
| LMArena Russian | — | 1392 |
| LMArena Spanish | — | 1416 |

## Instruction Following

- Codestral: —
- GPT-6 Luna: 74.7 (#85)

| Benchmark | Codestral | GPT-6 Luna |
|---|---|---|
| LMArena Instruction Following | — | 1414 |

## Long Context

- Codestral: —
- GPT-6 Luna: 43.3 (#95)

| Benchmark | Codestral | GPT-6 Luna |
|---|---|---|
| LMArena Longer Query | — | 1417 |

## Writing & Preference

- Codestral: —
- GPT-6 Luna: 58.4 (#115)

| Benchmark | Codestral | GPT-6 Luna |
|---|---|---|
| LMArena Text | — | 1392 |
| LMArena Creative Writing | — | 1357 |
| LMArena Multi-Turn | — | 1400 |

## FAQ

### Is Codestral better than GPT-6 Luna?

GPT-6 Luna is the stronger model overall, scoring 45.1 to 30.1 on the Noometry Index.

### Which is cheaper, Codestral or GPT-6 Luna?

GPT-6 Luna is cheaper. It lists at $0.10 per million input tokens and $0.50 per million output tokens; Codestral lists at $0.30 and $0.90.

### Is Codestral or GPT-6 Luna better for coding?

GPT-6 Luna scores higher on coding benchmarks: 54.7 versus 26.9 in the Noometry coding category.

### Which has the bigger context window?

GPT-6 Luna does, with 1.05M tokens against 256K.

### How many benchmarks do Codestral and GPT-6 Luna share?

0 benchmarks have published results for both models. Codestral has 7 scored results on Noometry and GPT-6 Luna has 18.