# Llama 4 Scout vs Qwen3 Max

> Qwen3 Max is the stronger model overall, scoring 43.5 to 27.4 on the Noometry Index. Llama 4 Scout costs 16× less per token, which makes it the better buy when Qwen3 Max's lead doesn't matter for your workload.

- Canonical page: https://noometry.com/compare/llama-4-scout-vs-qwen3-max
- Last updated: 2026-10-10
- Shared benchmarks: 25

## Summary

- They share 25 benchmarks with published results for both. Llama 4 Scout scores higher in 0 categories and Qwen3 Max in 8 categories; 8 gaps are clear of the uncertainty.
- The widest gap is in writing & preference, where Qwen3 Max leads 62.6 to 37.1.
- The biggest single-benchmark swing is OTIS Mock AIME 2024-2025: 7.8% for Llama 4 Scout and 73.3% for Qwen3 Max.
- Llama 4 Scout is cheaper at $0.10 / $0.30 per million input/output tokens, against $1.20 / $6 for Qwen3 Max.
- Qwen3 Max accepts more context: 262K tokens versus 128K.
- Llama 4 Scout has downloadable open weights; the other is API-only.

## Snapshot

| | Llama 4 Scout | Qwen3 Max |
|---|---|---|
| Provider | Meta | Alibaba (Qwen) |
| Noometry Index | 27.4 | 43.5 |
| Rank | 323 | 82 |
| Context | 128K | 262K |
| Input $/M | $0.10 | $1.20 |
| Output $/M | $0.30 | $6 |
| Weights | Open | Proprietary |

## Coding

- Llama 4 Scout: 19.9 (#334)
- Qwen3 Max: 42.4 (#93)

| Benchmark | Llama 4 Scout | Qwen3 Max |
|---|---|---|
| LMArena Coding | 1286 | 1456 |
| SWE-bench Verified (bash only) | 9.1% | — |
| SciCode | 17% | — |
| BigCodeBench Complete | 43.1% | — |
| ALE-Bench | — | 370.45 |

## Agentic & Tool Use

- Llama 4 Scout: 24.4 (#109)
- Qwen3 Max: —

| Benchmark | Llama 4 Scout | Qwen3 Max |
|---|---|---|
| Berkeley Function Calling Leaderboard | 28.1% | — |
| Vending-Bench 2 | — | 71.56 |

## Reasoning

- Llama 4 Scout: 9.2 (#339)
- Qwen3 Max: 22.0 (#186)

| Benchmark | Llama 4 Scout | Qwen3 Max |
|---|---|---|
| Kagi LLM Benchmark | 36.9% | 72.5% |
| LMArena Hard Prompts | 1266 | 1448 |
| DTBench | 57.9% | 82.1% |
| LMCA | 12% | 28.3% |
| Epoch Capabilities Index | 129.64 | 142.41 |
| ARC-AGI-2 | 0% | — |
| NYT Connections (extended) | — | 30.1% |
| ARC-AGI-1 | 0.5% | — |
| CritPt | 0% | — |
| Chess Puzzles | — | 4% |
| Mystery Game Puzzles | — | 5% |
| ForecastBench | 57.5 | — |

## Math

- Llama 4 Scout: 18.4 (#280)
- Qwen3 Max: 37.6 (#94)

| Benchmark | Llama 4 Scout | Qwen3 Max |
|---|---|---|
| OTIS Mock AIME 2024-2025 | 7.8% | 73.3% |
| LMArena Math | 1286 | 1447 |
| MATH Level 5 | 62.3% | 97.1% |
| FrontierMath (Tiers 1-3) | — | 18.9% |
| Omni-MATH | 37.3% | — |
| FrontierMath (Feb 2025 set) | 0% | — |

## Knowledge

- Llama 4 Scout: 31.8 (#208)
- Qwen3 Max: 48.1 (#72)

| Benchmark | Llama 4 Scout | Qwen3 Max |
|---|---|---|
| GPQA Diamond | 51.8% | 72.6% |
| LMArena Expert | 1236 | 1457 |
| SimpleQA Verified | — | 48.7% |
| MMLU-Pro | 74.2% | — |
| Vectara Hallucination Rate | 7.7% | — |
| GPQA (HELM) | 50.7% | — |

## Multimodal

- Llama 4 Scout: 31.6 (#98)
- Qwen3 Max: —

| Benchmark | Llama 4 Scout | Qwen3 Max |
|---|---|---|
| LMArena Vision | 1118 | — |
| SpatialViz-Bench | 34.2% | — |

## Multilingual

- Llama 4 Scout: 41.3 (#207)
- Qwen3 Max: 53.9 (#58)

| Benchmark | Llama 4 Scout | Qwen3 Max |
|---|---|---|
| LMArena Non-English | 1253 | 1429 |
| LMArena Chinese | 1256 | 1479 |
| LMArena French | 1283 | 1450 |
| LMArena German | 1274 | 1463 |
| LMArena Japanese | 1209 | 1399 |
| LMArena Korean | 1207 | 1397 |
| LMArena Russian | 1263 | 1428 |
| LMArena Spanish | 1277 | 1463 |

## Instruction Following

- Llama 4 Scout: 66.0 (#212)
- Qwen3 Max: 75.0 (#81)

| Benchmark | Llama 4 Scout | Qwen3 Max |
|---|---|---|
| LMArena Instruction Following | 1248 | 1420 |
| IFEval | 81.8% | — |

## Long Context

- Llama 4 Scout: 27.5 (#289)
- Qwen3 Max: 41.6 (#129)

| Benchmark | Llama 4 Scout | Qwen3 Max |
|---|---|---|
| Fiction.LiveBench | 36% | 66.7% |
| LMArena Longer Query | 1265 | 1439 |
| CL-bench | — | 14.5% |

## Writing & Preference

- Llama 4 Scout: 37.1 (#256)
- Qwen3 Max: 62.6 (#72)

| Benchmark | Llama 4 Scout | Qwen3 Max |
|---|---|---|
| LMArena Text | 1279 | 1439 |
| LMArena Creative Writing | 1249 | 1401 |
| LMArena Multi-Turn | 1281 | 1446 |
| EQ-Bench Creative Writing | 783 | — |
| WildBench | 78% | — |

## FAQ

### Is Llama 4 Scout better than Qwen3 Max?

Qwen3 Max is the stronger model overall, scoring 43.5 to 27.4 on the Noometry Index. Llama 4 Scout costs 16× less per token, which makes it the better buy when Qwen3 Max's lead doesn't matter for your workload.

### Which is cheaper, Llama 4 Scout or Qwen3 Max?

Llama 4 Scout is cheaper. It lists at $0.10 per million input tokens and $0.30 per million output tokens; Qwen3 Max lists at $1.20 and $6.

### Is Llama 4 Scout or Qwen3 Max better for coding?

Qwen3 Max scores higher on coding benchmarks: 42.4 versus 19.9 in the Noometry coding category.

### Which has the bigger context window?

Qwen3 Max does, with 262K tokens against 128K.

### How many benchmarks do Llama 4 Scout and Qwen3 Max share?

25 benchmarks have published results for both models. Llama 4 Scout has 43 scored results on Noometry and Qwen3 Max has 33.