# GPT-5.1 vs Qwen3-4B

> GPT-5.1 is the stronger model overall, scoring 48.8 to 31.5 on the Noometry Index.

- Canonical page: https://noometry.com/compare/gpt-5-1-vs-qwen3-4b
- Last updated: 2026-10-10
- Shared benchmarks: 4

## Summary

- They share 4 benchmarks with published results for both. GPT-5.1 scores higher in 4 categories and Qwen3-4B in 0 categories; 4 gaps are clear of the uncertainty.
- The widest gap is in math, where GPT-5.1 leads 50.3 to 28.5.
- The biggest single-benchmark swing is OTIS Mock AIME 2024-2025: 88.6% for GPT-5.1 and 52.2% for Qwen3-4B.
- Qwen3-4B has downloadable open weights; the other is API-only.

## Snapshot

| | GPT-5.1 | Qwen3-4B |
|---|---|---|
| Provider | OpenAI | Alibaba (Qwen) |
| Noometry Index | 48.8 | 31.5 |
| Rank | 47 | 261 |
| Context | 400K | — |
| Input $/M | $1.25 | — |
| Output $/M | $10 | — |
| Weights | Proprietary | Open |

## Coding

- GPT-5.1: 46.9 (#59)
- Qwen3-4B: —

| Benchmark | GPT-5.1 | Qwen3-4B |
|---|---|---|
| SWE-bench Verified | 68% | — |
| SWE-bench Verified (bash only) | 66% | — |
| LMArena WebDev | 1395 | — |
| SciCode | 43.3% | — |
| GSO | 13.7% | — |
| WeirdML | 60.8% | — |
| LiveBench Coding | 72.5% | — |
| LMArena Coding | 1454 | — |
| ALE-Bench | 1,192 | — |

## Agentic & Tool Use

- GPT-5.1: 32.5 (#55)
- Qwen3-4B: 27.3 (#94)

| Benchmark | GPT-5.1 | Qwen3-4B |
|---|---|---|
| Terminal-Bench | 47.6% | — |
| Berkeley Function Calling Leaderboard | — | 35.7% |
| DeepResearch Bench | 42.8% | — |
| LMArena Search | 1199 | — |
| Vending-Bench 2 | 1,473 | — |

## Reasoning

- GPT-5.1: 39.9 (#50)
- Qwen3-4B: 18.5 (#264)

| Benchmark | GPT-5.1 | Qwen3-4B |
|---|---|---|
| Chess Puzzles | 32% | 4% |
| ARC-AGI-2 | 17.6% | — |
| SimpleBench | 53.2% | — |
| ARC-AGI-1 | 72.8% | — |
| CritPt | 4.9% | — |
| EnigmaEval | 11.2% | — |
| LiveBench Reasoning | 95.8% | — |
| LMArena Hard Prompts | 1457 | — |
| Mystery Game Puzzles | 19% | — |
| DTBench | 90.1% | — |
| LiveBench Data Analysis | 72.1% | — |
| LMCA | 43.9% | — |
| Epoch Capabilities Index | 149.66 | — |
| ForecastBench | 58.1 | — |
| LiveBench | 78.8% | — |

## Math

- GPT-5.1: 50.3 (#45)
- Qwen3-4B: 28.5 (#228)

| Benchmark | GPT-5.1 | Qwen3-4B |
|---|---|---|
| OTIS Mock AIME 2024-2025 | 88.6% | 52.2% |
| MathArena Final-Answer Competitions | — | 38.5% |
| Omni-MATH | 46.4% | — |
| LiveBench Math | 94.5% | — |
| LMArena Math | 1446 | — |
| FrontierMath (Feb 2025 set) | 31% | — |
| FrontierMath Tier 4 (v1) | 12.5% | — |

## Knowledge

- GPT-5.1: 50.5 (#65)
- Qwen3-4B: 33.0 (#198)

| Benchmark | GPT-5.1 | Qwen3-4B |
|---|---|---|
| GPQA Diamond | 87.6% | 52.3% |
| Vectara Hallucination Rate | 10.9% | 5.7% |
| Humanity's Last Exam | 23.7% | — |
| SimpleQA Verified | 48% | — |
| MMLU-Pro | 57.9% | — |
| GPQA (HELM) | 44.2% | — |
| LMArena Expert | 1470 | — |

## Multimodal

- GPT-5.1: 44.0 (#17)
- Qwen3-4B: —

| Benchmark | GPT-5.1 | Qwen3-4B |
|---|---|---|
| LMArena Vision | 1250 | — |
| VPCT | 58.7% | — |
| LMArena Document | 1403 | — |

## Multilingual

- GPT-5.1: 54.0 (#51)
- Qwen3-4B: —

| Benchmark | GPT-5.1 | Qwen3-4B |
|---|---|---|
| LMArena Non-English | 1431 | — |
| LMArena Chinese | 1495 | — |
| LMArena French | 1451 | — |
| LMArena German | 1439 | — |
| LMArena Japanese | 1453 | — |
| LMArena Korean | 1401 | — |
| LMArena Russian | 1435 | — |
| LMArena Spanish | 1430 | — |

## Instruction Following

- GPT-5.1: 84.0 (#1)
- Qwen3-4B: —

| Benchmark | GPT-5.1 | Qwen3-4B |
|---|---|---|
| LiveBench Instruction Following | 93.3% | — |
| IFEval | 93.5% | — |
| LMArena Instruction Following | 1442 | — |

## Long Context

- GPT-5.1: 47.6 (#14)
- Qwen3-4B: —

| Benchmark | GPT-5.1 | Qwen3-4B |
|---|---|---|
| CL-bench | 23.7% | — |
| CL-bench Life | 17.3% | — |
| LMArena Longer Query | 1448 | — |

## Writing & Preference

- GPT-5.1: 64.7 (#53)
- Qwen3-4B: —

| Benchmark | GPT-5.1 | Qwen3-4B |
|---|---|---|
| LMArena Text | 1443 | — |
| LMArena Creative Writing | 1427 | — |
| WildBench | 86.3% | — |
| LMArena Multi-Turn | 1449 | — |
| LiveBench Language | 80.2% | — |

## FAQ

### Is GPT-5.1 better than Qwen3-4B?

GPT-5.1 is the stronger model overall, scoring 48.8 to 31.5 on the Noometry Index.

### How many benchmarks do GPT-5.1 and Qwen3-4B share?

4 benchmarks have published results for both models. GPT-5.1 has 63 scored results on Noometry and Qwen3-4B has 6.