# Kimi K3 vs Qwen2.5 7B Instruct

> Kimi K3 is the stronger model overall, scoring 59.5 to 29.0 on the Noometry Index. Qwen2.5 7B Instruct costs 20× less per token, which makes it the better buy when Kimi K3's lead doesn't matter for your workload.

- Canonical page: https://noometry.com/compare/kimi-k3-vs-qwen2-5-7b-instruct
- Last updated: 2026-10-10
- Shared benchmarks: 6

## Summary

- They share 6 benchmarks with published results for both. Kimi K3 scores higher in 7 categories and Qwen2.5 7B Instruct in 0 categories; 7 gaps are clear of the uncertainty.
- The widest gap is in math, where Kimi K3 leads 74.2 to 12.6.
- The biggest single-benchmark swing is OTIS Mock AIME 2024-2025: 97.2% for Kimi K3 and 2.5% for Qwen2.5 7B Instruct.
- Qwen2.5 7B Instruct is cheaper at $0.17 / $0.70 per million input/output tokens, against $3 / $15 for Kimi K3.
- Kimi K3 accepts more context: 1.05M tokens versus 131K.

## Snapshot

| | Kimi K3 | Qwen2.5 7B Instruct |
|---|---|---|
| Provider | Moonshot AI | Alibaba (Qwen) |
| Noometry Index | 59.5 | 29.0 |
| Rank | 15 | 320 |
| Context | 1.05M | 131K |
| Input $/M | $3 | $0.17 |
| Output $/M | $15 | $0.70 |
| Weights | Open | Open |

## Coding

- Kimi K3: 61.0 (#10)
- Qwen2.5 7B Instruct: 36.5 (#208)

| Benchmark | Kimi K3 | Qwen2.5 7B Instruct |
|---|---|---|
| DeepSWE | 68.5% | — |
| FrontierCode | 44.2% | — |
| LMArena WebDev | 1654 | — |
| FrontierSWE | 25.9% | — |
| SciCode | 59.5% | — |
| WeirdML | 82.6% | — |
| BigCodeBench Instruct | — | 37.6% |
| LMArena Coding | 1508 | — |
| BigCodeBench Complete | — | 46.1% |
| ALE-Bench | 1,524 | — |

## Agentic & Tool Use

- Kimi K3: 41.8 (#20)
- Qwen2.5 7B Instruct: 23.8 (#124)

| Benchmark | Kimi K3 | Qwen2.5 7B Instruct |
|---|---|---|
| APEX-Agents | 50.6% | — |
| τ²-bench Banking | 37.1% | — |
| PostTrainBench | 32% | — |
| BALROG | — | 7.8% |
| GBAEval | 48.3% | — |
| GDP.pdf | 19% | — |
| Vending-Bench 2 | 5,165 | — |

## Reasoning

- Kimi K3: 63.0 (#17)
- Qwen2.5 7B Instruct: 14.8 (#322)

| Benchmark | Kimi K3 | Qwen2.5 7B Instruct |
|---|---|---|
| Chess Puzzles | 39% | 0% |
| DTBench | 91.2% | 47.7% |
| LMCA | 52.7% | 6.4% |
| Epoch Capabilities Index | 157.45 | 118.51 |
| ARC-AGI-2 | 60.4% | — |
| SimpleBench | 60.7% | — |
| NYT Connections (extended) | 93.6% | — |
| ARC-AGI-1 | 94.5% | — |
| CritPt | 23.4% | — |
| LMArena Hard Prompts | 1496 | — |
| Mystery Game Puzzles | 26% | — |
| Surface Evolver Bench | 95% | — |
| ForecastBench | 61.1 | — |

## Math

- Kimi K3: 74.2 (#16)
- Qwen2.5 7B Instruct: 12.6 (#306)

| Benchmark | Kimi K3 | Qwen2.5 7B Instruct |
|---|---|---|
| OTIS Mock AIME 2024-2025 | 97.2% | 2.5% |
| FrontierMath (Tiers 1-3) | 72.2% | — |
| FrontierMath Tier 4 | 39% | — |
| MathArena Final-Answer Competitions | 87.8% | — |
| ProofBench | 87% | — |
| Omni-MATH | — | 29.4% |
| LMArena Math | 1491 | — |

## Knowledge

- Kimi K3: 63.2 (#21)
- Qwen2.5 7B Instruct: 17.0 (#286)

| Benchmark | Kimi K3 | Qwen2.5 7B Instruct |
|---|---|---|
| GPQA Diamond | 93.1% | 35.5% |
| SimpleQA Verified | 50.6% | — |
| MMLU-Pro | — | 53.9% |
| GPQA (HELM) | — | 34.1% |
| LMArena Expert | 1521 | — |
| MMLU | — | 72.9% |

## Multimodal

- Kimi K3: 37.8 (#70)
- Qwen2.5 7B Instruct: —

| Benchmark | Kimi K3 | Qwen2.5 7B Instruct |
|---|---|---|
| Blueprint-Bench 2 | 29.5% | — |
| Furniture Assembly | 34.2% | — |

## Multilingual

- Kimi K3: 56.3 (#21)
- Qwen2.5 7B Instruct: —

| Benchmark | Kimi K3 | Qwen2.5 7B Instruct |
|---|---|---|
| LMArena Non-English | 1466 | — |
| LMArena Chinese | 1529 | — |
| LMArena French | 1491 | — |
| LMArena German | 1488 | — |
| LMArena Japanese | 1487 | — |
| LMArena Korean | 1458 | — |
| LMArena Russian | 1482 | — |
| LMArena Spanish | 1472 | — |

## Instruction Following

- Kimi K3: 77.7 (#14)
- Qwen2.5 7B Instruct: 63.2 (#231)

| Benchmark | Kimi K3 | Qwen2.5 7B Instruct |
|---|---|---|
| IFEval | — | 74.1% |
| LMArena Instruction Following | 1483 | — |

## Long Context

- Kimi K3: 45.8 (#29)
- Qwen2.5 7B Instruct: —

| Benchmark | Kimi K3 | Qwen2.5 7B Instruct |
|---|---|---|
| LMArena Longer Query | 1494 | — |

## Writing & Preference

- Kimi K3: 76.6 (#4)
- Qwen2.5 7B Instruct: 48.8 (#195)

| Benchmark | Kimi K3 | Qwen2.5 7B Instruct |
|---|---|---|
| LMArena Text | 1476 | — |
| LMArena Creative Writing | 1454 | — |
| EQ-Bench Creative Writing | 2082 | — |
| WildBench | — | 73.1% |
| EQ-Bench 4 | 1339 | — |
| LMArena Multi-Turn | 1488 | — |

## FAQ

### Is Kimi K3 better than Qwen2.5 7B Instruct?

Kimi K3 is the stronger model overall, scoring 59.5 to 29.0 on the Noometry Index. Qwen2.5 7B Instruct costs 20× less per token, which makes it the better buy when Kimi K3's lead doesn't matter for your workload.

### Which is cheaper, Kimi K3 or Qwen2.5 7B Instruct?

Qwen2.5 7B Instruct is cheaper. It lists at $0.17 per million input tokens and $0.70 per million output tokens; Kimi K3 lists at $3 and $15.

### Is Kimi K3 or Qwen2.5 7B Instruct better for coding?

Kimi K3 scores higher on coding benchmarks: 61.0 versus 36.5 in the Noometry coding category.

### Which has the bigger context window?

Kimi K3 does, with 1.05M tokens against 131K.

### How many benchmarks do Kimi K3 and Qwen2.5 7B Instruct share?

6 benchmarks have published results for both models. Kimi K3 has 53 scored results on Noometry and Qwen2.5 7B Instruct has 15.