# Codestral vs Qwen3.8 27B

> Qwen3.8 27B is the stronger model overall, scoring 46.0 to 30.6 on the Noometry Index.

- Canonical page: https://noometry.com/compare/codestral-vs-qwen3-8-27b
- Last updated: 2026-10-12
- Shared benchmarks: 0

## Summary

- The widest gap is in coding, where Qwen3.8 27B leads 50.5 to 27.3.
- Codestral is cheaper at $0.30 / $0.90 per million input/output tokens, against $0.04 / $2.30 for Qwen3.8 27B.
- Qwen3.8 27B accepts more context: 262K tokens versus 256K.
- Qwen3.8 27B has downloadable open weights; the other is API-only.

## Snapshot

| | Codestral | Qwen3.8 27B |
|---|---|---|
| Provider | Mistral AI | Alibaba (Qwen) |
| Noometry Index | 30.6 | 46.0 |
| Rank | 290 | 68 |
| Context | 256K | 262K |
| Input $/M | $0.30 | $0.04 |
| Output $/M | $0.90 | $2.30 |
| Weights | Proprietary | Open |

## Coding

- Codestral: 27.3 (#321)
- Qwen3.8 27B: 50.5 (#44)

| Benchmark | Codestral | Qwen3.8 27B |
|---|---|---|
| Aider Polyglot | 11.1% | — |
| LMArena WebDev | — | 1593 |
| SciCode | — | 46.6% |
| BigCodeBench Instruct | 41.8% | — |
| LMArena Coding | — | 1482 |
| BigCodeBench Complete | 52.5% | — |
| ALE-Bench | 137.78 | — |
| HumanEval+ | 73.8% | — |
| MBPP+ | 61.9% | — |

## Agentic & Tool Use

- Codestral: —
- Qwen3.8 27B: 32.9 (#57)

| Benchmark | Codestral | Qwen3.8 27B |
|---|---|---|
| APEX-Agents | — | 47.5% |

## Reasoning

- Codestral: 19.8 (#251)
- Qwen3.8 27B: 41.0 (#54)

| Benchmark | Codestral | Qwen3.8 27B |
|---|---|---|
| ARC-AGI-2 | — | 42.4% |
| Kagi LLM Benchmark | 32.5% | — |
| NYT Connections (extended) | — | 54.5% |
| ARC-AGI-1 | — | 87.5% |
| CritPt | — | 5.4% |
| LMArena Hard Prompts | — | 1460 |
| DTBench | — | 88% |
| LMCA | — | 41.4% |
| Surface Evolver Bench | — | 45% |
| Epoch Capabilities Index | — | 149.38 |

## Math

- Codestral: —
- Qwen3.8 27B: 37.1 (#161)

| Benchmark | Codestral | Qwen3.8 27B |
|---|---|---|
| ProofBench | — | 16% |
| LMArena Math | — | 1456 |

## Knowledge

- Codestral: —
- Qwen3.8 27B: 41.6 (#109)

| Benchmark | Codestral | Qwen3.8 27B |
|---|---|---|
| LMArena Expert | — | 1482 |

## Multimodal

- Codestral: —
- Qwen3.8 27B: 41.3 (#37)

| Benchmark | Codestral | Qwen3.8 27B |
|---|---|---|
| LMArena Vision | — | 1271 |

## Multilingual

- Codestral: —
- Qwen3.8 27B: 53.7 (#60)

| Benchmark | Codestral | Qwen3.8 27B |
|---|---|---|
| LMArena Non-English | — | 1430 |
| LMArena Chinese | — | 1504 |
| LMArena French | — | 1465 |
| LMArena German | — | 1438 |
| LMArena Japanese | — | 1384 |
| LMArena Korean | — | 1393 |
| LMArena Russian | — | 1415 |
| LMArena Spanish | — | 1448 |

## Instruction Following

- Codestral: —
- Qwen3.8 27B: 75.8 (#53)

| Benchmark | Codestral | Qwen3.8 27B |
|---|---|---|
| LMArena Instruction Following | — | 1439 |

## Long Context

- Codestral: —
- Qwen3.8 27B: 44.3 (#70)

| Benchmark | Codestral | Qwen3.8 27B |
|---|---|---|
| LMArena Longer Query | — | 1450 |

## Writing & Preference

- Codestral: —
- Qwen3.8 27B: 65.8 (#43)

| Benchmark | Codestral | Qwen3.8 27B |
|---|---|---|
| LMArena Text | — | 1441 |
| LMArena Creative Writing | — | 1384 |
| EQ-Bench Creative Writing | — | 1671 |
| LMArena Multi-Turn | — | 1441 |

## FAQ

### Is Codestral better than Qwen3.8 27B?

Qwen3.8 27B is the stronger model overall, scoring 46.0 to 30.6 on the Noometry Index.

### Which is cheaper, Codestral or Qwen3.8 27B?

Codestral is cheaper. It lists at $0.30 per million input tokens and $0.90 per million output tokens; Qwen3.8 27B lists at $0.04 and $2.30.

### Is Codestral or Qwen3.8 27B better for coding?

Qwen3.8 27B scores higher on coding benchmarks: 50.5 versus 27.3 in the Noometry coding category.

### Which has the bigger context window?

Qwen3.8 27B does, with 262K tokens against 256K.

### How many benchmarks do Codestral and Qwen3.8 27B share?

0 benchmarks have published results for both models. Codestral has 7 scored results on Noometry and Qwen3.8 27B has 31.