# Kimi K2.5 vs Qwen3-4B

> Kimi K2.5 is the stronger model overall, scoring 47.8 to 31.5 on the Noometry Index.

- Canonical page: https://noometry.com/compare/kimi-k2-5-vs-qwen3-4b
- Last updated: 2026-10-10
- Shared benchmarks: 5

## Summary

- They share 5 benchmarks with published results for both. Kimi K2.5 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 Kimi K2.5 leads 50.3 to 28.5.
- The biggest single-benchmark swing is OTIS Mock AIME 2024-2025: 92.2% for Kimi K2.5 and 52.2% for Qwen3-4B.

## Snapshot

| | Kimi K2.5 | Qwen3-4B |
|---|---|---|
| Provider | Moonshot AI | Alibaba (Qwen) |
| Noometry Index | 47.8 | 31.5 |
| Rank | 50 | 261 |
| Context | 262K | — |
| Input $/M | $0.45 | — |
| Output $/M | $2.25 | — |
| Weights | Open | Open |

## Coding

- Kimi K2.5: 48.7 (#48)
- Qwen3-4B: —

| Benchmark | Kimi K2.5 | Qwen3-4B |
|---|---|---|
| SWE-bench Verified | 73.8% | — |
| SWE-bench Verified (bash only) | 70.8% | — |
| LMArena WebDev | 1436 | — |
| SWE-bench Multilingual | 67.3% | — |
| SciCode | 49% | — |
| WeirdML | 45.6% | — |
| LMArena Coding | 1474 | — |
| ALE-Bench | 821.65 | — |

## Agentic & Tool Use

- Kimi K2.5: 34.0 (#44)
- Qwen3-4B: 27.3 (#94)

| Benchmark | Kimi K2.5 | Qwen3-4B |
|---|---|---|
| Terminal-Bench | 43.2% | — |
| Berkeley Function Calling Leaderboard | — | 35.7% |
| OSWorld | 63.3% | — |
| Vending-Bench 2 | 1,198 | — |

## Reasoning

- Kimi K2.5: 31.2 (#74)
- Qwen3-4B: 18.5 (#264)

| Benchmark | Kimi K2.5 | Qwen3-4B |
|---|---|---|
| Chess Puzzles | 12% | 4% |
| ARC-AGI-2 | 11.8% | — |
| SimpleBench | 46.8% | — |
| Kagi LLM Benchmark | 78.5% | — |
| NYT Connections (extended) | 69.9% | — |
| ARC-AGI-1 | 65.3% | — |
| CritPt | 3.1% | — |
| EnigmaEval | 3.4% | — |
| Thematic Generalization | 69.4% | — |
| LMArena Hard Prompts | 1454 | — |
| Epoch Capabilities Index | 148.05 | — |

## Math

- Kimi K2.5: 50.3 (#46)
- Qwen3-4B: 28.5 (#228)

| Benchmark | Kimi K2.5 | Qwen3-4B |
|---|---|---|
| MathArena Final-Answer Competitions | 62.3% | 38.5% |
| OTIS Mock AIME 2024-2025 | 92.2% | 52.2% |
| LMArena Math | 1470 | — |
| FrontierMath (Feb 2025 set) | 27.9% | — |
| FrontierMath Tier 4 (v1) | 4.2% | — |

## Knowledge

- Kimi K2.5: 53.6 (#52)
- Qwen3-4B: 33.0 (#198)

| Benchmark | Kimi K2.5 | Qwen3-4B |
|---|---|---|
| GPQA Diamond | 87.6% | 52.3% |
| Vectara Hallucination Rate | 14.2% | 5.7% |
| Humanity's Last Exam | 24.4% | — |
| SimpleQA Verified | 34.3% | — |
| LMArena Expert | 1466 | — |

## Multimodal

- Kimi K2.5: 40.4 (#37)
- Qwen3-4B: —

| Benchmark | Kimi K2.5 | Qwen3-4B |
|---|---|---|
| LMArena Vision | 1269 | — |
| LMArena Document | 1430 | — |

## Multilingual

- Kimi K2.5: 54.1 (#50)
- Qwen3-4B: —

| Benchmark | Kimi K2.5 | Qwen3-4B |
|---|---|---|
| LMArena Non-English | 1433 | — |
| LMArena Chinese | 1495 | — |
| LMArena French | 1454 | — |
| LMArena German | 1441 | — |
| LMArena Japanese | 1422 | — |
| LMArena Korean | 1412 | — |
| LMArena Russian | 1436 | — |
| LMArena Spanish | 1450 | — |

## Instruction Following

- Kimi K2.5: 75.5 (#60)
- Qwen3-4B: —

| Benchmark | Kimi K2.5 | Qwen3-4B |
|---|---|---|
| LMArena Instruction Following | 1431 | — |

## Long Context

- Kimi K2.5: 52.1 (#7)
- Qwen3-4B: —

| Benchmark | Kimi K2.5 | Qwen3-4B |
|---|---|---|
| Fiction.LiveBench | 86.1% | — |
| CL-bench | 19.3% | — |
| CL-bench Life | 13.2% | — |
| LMArena Longer Query | 1445 | — |

## Writing & Preference

- Kimi K2.5: 65.3 (#51)
- Qwen3-4B: —

| Benchmark | Kimi K2.5 | Qwen3-4B |
|---|---|---|
| LMArena Text | 1445 | — |
| LMArena Creative Writing | 1423 | — |
| EQ-Bench Creative Writing | 1579 | — |
| LMArena Multi-Turn | 1444 | — |

## FAQ

### Is Kimi K2.5 better than Qwen3-4B?

Kimi K2.5 is the stronger model overall, scoring 47.8 to 31.5 on the Noometry Index.

### How many benchmarks do Kimi K2.5 and Qwen3-4B share?

5 benchmarks have published results for both models. Kimi K2.5 has 51 scored results on Noometry and Qwen3-4B has 6.