# GPT-4 vs Qwen3-30B-A3B

> Qwen3-30B-A3B is the stronger model overall, scoring 38.9 to 29.1 on the Noometry Index.

- Canonical page: https://noometry.com/compare/gpt-4-vs-qwen3-30b-a3b
- Last updated: 2026-10-12
- Shared benchmarks: 24

## Summary

- They share 24 benchmarks with published results for both. GPT-4 scores higher in 1 category and Qwen3-30B-A3B in 7 categories; 8 gaps are clear of the uncertainty.
- The widest gap is in math, where Qwen3-30B-A3B leads 37.4 to 10.8.
- The biggest single-benchmark swing is OTIS Mock AIME 2024-2025: 1.1% for GPT-4 and 70.3% for Qwen3-30B-A3B.
- Qwen3-30B-A3B is cheaper at $0.12 / $0.50 per million input/output tokens, against $30 / $60 for GPT-4.
- Qwen3-30B-A3B accepts more context: 41K tokens versus 8K.
- Qwen3-30B-A3B has downloadable open weights; the other is API-only.

## Snapshot

| | GPT-4 | Qwen3-30B-A3B |
|---|---|---|
| Provider | OpenAI | Alibaba (Qwen) |
| Noometry Index | 29.1 | 38.9 |
| Rank | 316 | 179 |
| Context | 8K | 41K |
| Input $/M | $30 | $0.12 |
| Output $/M | $60 | $0.50 |
| Weights | Proprietary | Open |

## Coding

- GPT-4: 31.6 (#283)
- Qwen3-30B-A3B: 37.5 (#194)

| Benchmark | GPT-4 | Qwen3-30B-A3B |
|---|---|---|
| WeirdML | 12.4% | 29.8% |
| LMArena Coding | 1254 | 1416 |
| SciCode | — | 33.3% |
| BigCodeBench Instruct | 46% | — |
| BigCodeBench Complete | 57.2% | — |
| HumanEval+ | 79.3% | — |

## Agentic & Tool Use

- GPT-4: —
- Qwen3-30B-A3B: 29.8 (#82)

| Benchmark | GPT-4 | Qwen3-30B-A3B |
|---|---|---|
| Berkeley Function Calling Leaderboard | — | 41.4% |
| METR Time Horizons | 36.1% | — |

## Reasoning

- GPT-4: 17.8 (#289)
- Qwen3-30B-A3B: 22.2 (#204)

| Benchmark | GPT-4 | Qwen3-30B-A3B |
|---|---|---|
| Chess Puzzles | 4% | 8% |
| LMArena Hard Prompts | 1241 | 1398 |
| DTBench | 62.7% | 69.3% |
| LMCA | 17.1% | 22.4% |
| Epoch Capabilities Index | 125.89 | 139.63 |
| Kagi LLM Benchmark | — | 54.9% |
| CritPt | — | 0.3% |
| Mystery Game Puzzles | 12% | — |
| BIG-Bench Hard | 75.1% | — |
| ForecastBench | 57.8 | — |
| HellaSwag | 95.3% | — |
| WinoGrande | 87.5% | — |

## Math

- GPT-4: 10.8 (#309)
- Qwen3-30B-A3B: 37.4 (#157)

| Benchmark | GPT-4 | Qwen3-30B-A3B |
|---|---|---|
| OTIS Mock AIME 2024-2025 | 1.1% | 70.3% |
| LMArena Math | 1269 | 1394 |
| MathArena Final-Answer Competitions | — | 47.8% |
| MATH Level 5 | 23% | — |
| GSM8K | 92% | — |

## Knowledge

- GPT-4: 18.4 (#282)
- Qwen3-30B-A3B: 41.8 (#105)

| Benchmark | GPT-4 | Qwen3-30B-A3B |
|---|---|---|
| GPQA Diamond | 35.7% | 70.1% |
| LMArena Expert | 1211 | 1396 |
| Confabulations | — | 12.3% |
| MMLU | 86.4% | — |
| TriviaQA | 84.8% | — |

## Multilingual

- GPT-4: 40.6 (#215)
- Qwen3-30B-A3B: 49.5 (#132)

| Benchmark | GPT-4 | Qwen3-30B-A3B |
|---|---|---|
| LMArena Non-English | 1246 | 1372 |
| LMArena Chinese | 1242 | 1433 |
| LMArena French | 1283 | 1418 |
| LMArena German | 1251 | 1380 |
| LMArena Japanese | 1209 | 1337 |
| LMArena Korean | 1184 | 1331 |
| LMArena Russian | 1251 | 1370 |
| LMArena Spanish | 1261 | 1404 |

## Instruction Following

- GPT-4: 65.3 (#222)
- Qwen3-30B-A3B: 72.0 (#142)

| Benchmark | GPT-4 | Qwen3-30B-A3B |
|---|---|---|
| LMArena Instruction Following | 1241 | 1363 |

## Long Context

- GPT-4: 37.7 (#212)
- Qwen3-30B-A3B: 31.0 (#283)

| Benchmark | GPT-4 | Qwen3-30B-A3B |
|---|---|---|
| LMArena Longer Query | 1244 | 1379 |
| Fiction.LiveBench | — | 40.6% |

## Writing & Preference

- GPT-4: 34.9 (#268)
- Qwen3-30B-A3B: 55.6 (#143)

| Benchmark | GPT-4 | Qwen3-30B-A3B |
|---|---|---|
| LMArena Text | 1263 | 1384 |
| LMArena Creative Writing | 1244 | 1317 |
| LMArena Multi-Turn | 1257 | 1378 |
| Short-Story Creative Writing | — | 75.3% |
| EQ-Bench Creative Writing | 752 | — |

## FAQ

### Is GPT-4 better than Qwen3-30B-A3B?

Qwen3-30B-A3B is the stronger model overall, scoring 38.9 to 29.1 on the Noometry Index.

### Which is cheaper, GPT-4 or Qwen3-30B-A3B?

Qwen3-30B-A3B is cheaper. It lists at $0.12 per million input tokens and $0.50 per million output tokens; GPT-4 lists at $30 and $60.

### Is GPT-4 or Qwen3-30B-A3B better for coding?

Qwen3-30B-A3B scores higher on coding benchmarks: 37.5 versus 31.6 in the Noometry coding category.

### Which has the bigger context window?

Qwen3-30B-A3B does, with 41K tokens against 8K.

### How many benchmarks do GPT-4 and Qwen3-30B-A3B share?

24 benchmarks have published results for both models. GPT-4 has 38 scored results on Noometry and Qwen3-30B-A3B has 32.