# Kimi K2.5 vs Llama 4 Scout

> Kimi K2.5 is the stronger model overall, scoring 47.8 to 27.4 on the Noometry Index. Llama 4 Scout costs 6.0× less per token, which makes it the better buy when Kimi K2.5's lead doesn't matter for your workload.

- Canonical page: https://noometry.com/compare/kimi-k2-5-vs-llama-4-scout
- Last updated: 2026-10-10
- Shared benchmarks: 31

## Summary

- They share 31 benchmarks with published results for both. Kimi K2.5 scores higher in 10 categories and Llama 4 Scout in 0 categories; 10 gaps are clear of the uncertainty.
- The widest gap is in math, where Kimi K2.5 leads 50.3 to 18.4.
- The biggest single-benchmark swing is OTIS Mock AIME 2024-2025: 92.2% for Kimi K2.5 and 7.8% for Llama 4 Scout.
- Llama 4 Scout is cheaper at $0.10 / $0.30 per million input/output tokens, against $0.45 / $2.25 for Kimi K2.5.
- Kimi K2.5 accepts more context: 262K tokens versus 128K.

## Snapshot

| | Kimi K2.5 | Llama 4 Scout |
|---|---|---|
| Provider | Moonshot AI | Meta |
| Noometry Index | 47.8 | 27.4 |
| Rank | 50 | 323 |
| Context | 262K | 128K |
| Input $/M | $0.45 | $0.10 |
| Output $/M | $2.25 | $0.30 |
| Weights | Open | Open |

## Coding

- Kimi K2.5: 48.7 (#48)
- Llama 4 Scout: 19.9 (#334)

| Benchmark | Kimi K2.5 | Llama 4 Scout |
|---|---|---|
| SWE-bench Verified (bash only) | 70.8% | 9.1% |
| SciCode | 49% | 17% |
| LMArena Coding | 1474 | 1286 |
| SWE-bench Verified | 73.8% | — |
| LMArena WebDev | 1436 | — |
| SWE-bench Multilingual | 67.3% | — |
| WeirdML | 45.6% | — |
| BigCodeBench Complete | — | 43.1% |
| ALE-Bench | 821.65 | — |

## Agentic & Tool Use

- Kimi K2.5: 34.0 (#44)
- Llama 4 Scout: 24.4 (#109)

| Benchmark | Kimi K2.5 | Llama 4 Scout |
|---|---|---|
| Terminal-Bench | 43.2% | — |
| Berkeley Function Calling Leaderboard | — | 28.1% |
| OSWorld | 63.3% | — |
| Vending-Bench 2 | 1,198 | — |

## Reasoning

- Kimi K2.5: 31.2 (#74)
- Llama 4 Scout: 9.2 (#339)

| Benchmark | Kimi K2.5 | Llama 4 Scout |
|---|---|---|
| ARC-AGI-2 | 11.8% | 0% |
| Kagi LLM Benchmark | 78.5% | 36.9% |
| ARC-AGI-1 | 65.3% | 0.5% |
| CritPt | 3.1% | 0% |
| LMArena Hard Prompts | 1454 | 1266 |
| Epoch Capabilities Index | 148.05 | 129.64 |
| SimpleBench | 46.8% | — |
| NYT Connections (extended) | 69.9% | — |
| Chess Puzzles | 12% | — |
| EnigmaEval | 3.4% | — |
| Thematic Generalization | 69.4% | — |
| DTBench | — | 57.9% |
| LMCA | — | 12% |
| ForecastBench | — | 57.5 |

## Math

- Kimi K2.5: 50.3 (#46)
- Llama 4 Scout: 18.4 (#280)

| Benchmark | Kimi K2.5 | Llama 4 Scout |
|---|---|---|
| OTIS Mock AIME 2024-2025 | 92.2% | 7.8% |
| LMArena Math | 1470 | 1286 |
| FrontierMath (Feb 2025 set) | 27.9% | 0% |
| MathArena Final-Answer Competitions | 62.3% | — |
| Omni-MATH | — | 37.3% |
| MATH Level 5 | — | 62.3% |
| FrontierMath Tier 4 (v1) | 4.2% | — |

## Knowledge

- Kimi K2.5: 53.6 (#52)
- Llama 4 Scout: 31.8 (#208)

| Benchmark | Kimi K2.5 | Llama 4 Scout |
|---|---|---|
| GPQA Diamond | 87.6% | 51.8% |
| Vectara Hallucination Rate | 14.2% | 7.7% |
| LMArena Expert | 1466 | 1236 |
| Humanity's Last Exam | 24.4% | — |
| SimpleQA Verified | 34.3% | — |
| MMLU-Pro | — | 74.2% |
| GPQA (HELM) | — | 50.7% |

## Multimodal

- Kimi K2.5: 40.4 (#37)
- Llama 4 Scout: 31.6 (#98)

| Benchmark | Kimi K2.5 | Llama 4 Scout |
|---|---|---|
| LMArena Vision | 1269 | 1118 |
| LMArena Document | 1430 | — |
| SpatialViz-Bench | — | 34.2% |

## Multilingual

- Kimi K2.5: 54.1 (#50)
- Llama 4 Scout: 41.3 (#207)

| Benchmark | Kimi K2.5 | Llama 4 Scout |
|---|---|---|
| LMArena Non-English | 1433 | 1253 |
| LMArena Chinese | 1495 | 1256 |
| LMArena French | 1454 | 1283 |
| LMArena German | 1441 | 1274 |
| LMArena Japanese | 1422 | 1209 |
| LMArena Korean | 1412 | 1207 |
| LMArena Russian | 1436 | 1263 |
| LMArena Spanish | 1450 | 1277 |

## Instruction Following

- Kimi K2.5: 75.5 (#60)
- Llama 4 Scout: 66.0 (#212)

| Benchmark | Kimi K2.5 | Llama 4 Scout |
|---|---|---|
| LMArena Instruction Following | 1431 | 1248 |
| IFEval | — | 81.8% |

## Long Context

- Kimi K2.5: 52.1 (#7)
- Llama 4 Scout: 27.5 (#289)

| Benchmark | Kimi K2.5 | Llama 4 Scout |
|---|---|---|
| Fiction.LiveBench | 86.1% | 36% |
| LMArena Longer Query | 1445 | 1265 |
| CL-bench | 19.3% | — |
| CL-bench Life | 13.2% | — |

## Writing & Preference

- Kimi K2.5: 65.3 (#51)
- Llama 4 Scout: 37.1 (#256)

| Benchmark | Kimi K2.5 | Llama 4 Scout |
|---|---|---|
| LMArena Text | 1445 | 1279 |
| LMArena Creative Writing | 1423 | 1249 |
| EQ-Bench Creative Writing | 1579 | 783 |
| LMArena Multi-Turn | 1444 | 1281 |
| WildBench | — | 78% |

## FAQ

### Is Kimi K2.5 better than Llama 4 Scout?

Kimi K2.5 is the stronger model overall, scoring 47.8 to 27.4 on the Noometry Index. Llama 4 Scout costs 6.0× less per token, which makes it the better buy when Kimi K2.5's lead doesn't matter for your workload.

### Which is cheaper, Kimi K2.5 or Llama 4 Scout?

Llama 4 Scout is cheaper. It lists at $0.10 per million input tokens and $0.30 per million output tokens; Kimi K2.5 lists at $0.45 and $2.25.

### Is Kimi K2.5 or Llama 4 Scout better for coding?

Kimi K2.5 scores higher on coding benchmarks: 48.7 versus 19.9 in the Noometry coding category.

### Which has the bigger context window?

Kimi K2.5 does, with 262K tokens against 128K.

### How many benchmarks do Kimi K2.5 and Llama 4 Scout share?

31 benchmarks have published results for both models. Kimi K2.5 has 51 scored results on Noometry and Llama 4 Scout has 43.