# GPT-5.1-Codex vs Qwen3.5 27B

> Qwen3.5 27B is the stronger model overall, scoring 41.9 to 38.6 on the Noometry Index.

- Canonical page: https://noometry.com/compare/gpt-5-1-codex-vs-qwen3-5-27b
- Last updated: 2026-10-12
- Shared benchmarks: 2

## Summary

- They share 2 benchmarks with published results for both. GPT-5.1-Codex scores higher in 1 category and Qwen3.5 27B in 1 category; 2 gaps are clear of the uncertainty.
- The widest gap is in math, where Qwen3.5 27B leads 38.8 to 30.3.
- Qwen3.5 27B is cheaper at $0.30 / $2.40 per million input/output tokens, against $1.25 / $10 for GPT-5.1-Codex.
- GPT-5.1-Codex accepts more context: 400K tokens versus 262K.
- Qwen3.5 27B has downloadable open weights; the other is API-only.

## Snapshot

| | GPT-5.1-Codex | Qwen3.5 27B |
|---|---|---|
| Provider | OpenAI | Alibaba (Qwen) |
| Noometry Index | 38.6 | 41.9 |
| Rank | 186 | 127 |
| Context | 400K | 262K |
| Input $/M | $1.25 | $0.30 |
| Output $/M | $10 | $2.40 |
| Weights | Proprietary | Open |

## Coding

- GPT-5.1-Codex: 41.9 (#116)
- Qwen3.5 27B: 38.9 (#168)

| Benchmark | GPT-5.1-Codex | Qwen3.5 27B |
|---|---|---|
| LMArena WebDev | 1337 | 1358 |
| ALE-Bench | 1,245 | 349.45 |
| SWE-bench Verified (bash only) | 66% | — |
| WeirdML | — | 39.5% |
| LMArena Coding | — | 1427 |

## Agentic & Tool Use

- GPT-5.1-Codex: 38.0 (#33)
- Qwen3.5 27B: —

| Benchmark | GPT-5.1-Codex | Qwen3.5 27B |
|---|---|---|
| Terminal-Bench | 60.4% | — |
| METR Time Horizons | 70.8% | — |
| Vending-Bench 2 | — | 201.98 |

## Reasoning

- GPT-5.1-Codex: —
- Qwen3.5 27B: 27.5 (#117)

| Benchmark | GPT-5.1-Codex | Qwen3.5 27B |
|---|---|---|
| NYT Connections (extended) | — | 47.9% |
| Thematic Generalization | — | 45.5% |
| LMArena Hard Prompts | — | 1414 |
| DTBench | — | 82.4% |
| LMCA | — | 34% |

## Math

- GPT-5.1-Codex: 30.3 (#235)
- Qwen3.5 27B: 38.8 (#127)

| Benchmark | GPT-5.1-Codex | Qwen3.5 27B |
|---|---|---|
| MathArena Final-Answer Competitions | — | 56.7% |
| ProofBench | 9% | — |
| LMArena Math | — | 1429 |

## Knowledge

- GPT-5.1-Codex: —
- Qwen3.5 27B: 38.0 (#150)

| Benchmark | GPT-5.1-Codex | Qwen3.5 27B |
|---|---|---|
| Vectara Hallucination Rate | — | 12.1% |
| LMArena Expert | — | 1428 |

## Multimodal

- GPT-5.1-Codex: —
- Qwen3.5 27B: 39.4 (#59)

| Benchmark | GPT-5.1-Codex | Qwen3.5 27B |
|---|---|---|
| LMArena Vision | — | 1241 |

## Multilingual

- GPT-5.1-Codex: —
- Qwen3.5 27B: 50.8 (#115)

| Benchmark | GPT-5.1-Codex | Qwen3.5 27B |
|---|---|---|
| LMArena Non-English | — | 1390 |
| LMArena Chinese | — | 1478 |
| LMArena French | — | 1410 |
| LMArena German | — | 1393 |
| LMArena Japanese | — | 1345 |
| LMArena Korean | — | 1358 |
| LMArena Russian | — | 1390 |
| LMArena Spanish | — | 1407 |

## Instruction Following

- GPT-5.1-Codex: —
- Qwen3.5 27B: 73.5 (#119)

| Benchmark | GPT-5.1-Codex | Qwen3.5 27B |
|---|---|---|
| LMArena Instruction Following | — | 1393 |

## Long Context

- GPT-5.1-Codex: —
- Qwen3.5 27B: 43.1 (#106)

| Benchmark | GPT-5.1-Codex | Qwen3.5 27B |
|---|---|---|
| LMArena Longer Query | — | 1413 |

## Writing & Preference

- GPT-5.1-Codex: —
- Qwen3.5 27B: 59.3 (#111)

| Benchmark | GPT-5.1-Codex | Qwen3.5 27B |
|---|---|---|
| LMArena Text | — | 1409 |
| LMArena Creative Writing | — | 1362 |
| LMArena Multi-Turn | — | 1410 |

## FAQ

### Is GPT-5.1-Codex better than Qwen3.5 27B?

Qwen3.5 27B is the stronger model overall, scoring 41.9 to 38.6 on the Noometry Index.

### Which is cheaper, GPT-5.1-Codex or Qwen3.5 27B?

Qwen3.5 27B is cheaper. It lists at $0.30 per million input tokens and $2.40 per million output tokens; GPT-5.1-Codex lists at $1.25 and $10.

### Is GPT-5.1-Codex or Qwen3.5 27B better for coding?

GPT-5.1-Codex scores higher on coding benchmarks: 41.9 versus 38.9 in the Noometry coding category.

### Which has the bigger context window?

GPT-5.1-Codex does, with 400K tokens against 262K.

### How many benchmarks do GPT-5.1-Codex and Qwen3.5 27B share?

2 benchmarks have published results for both models. GPT-5.1-Codex has 6 scored results on Noometry and Qwen3.5 27B has 28.