# Kimi K2 Benchmark Scores & Performance

> Kimi K2 by Moonshot AI scores 33.48/100 overall, ranking #168 out of 954 AI models.

Canonical page: https://benchlm.ai/models/kimi-k2

Last updated: October 17, 2026

## Model Details

| Property | Value |
|----------|-------|
| Creator | Moonshot AI |
| Source Type | Proprietary |
| Reasoning Type | Non-Reasoning |
| Context Window | 128K |
| Overall Score | 33.48/100 |
| Overall Rank | #168 of 954 |

## Family & Coverage

- Family: Kimi K2
- Variant: base
- Benchmarks covered: 24 of 639
- Coverage note: BenchLM currently has partial benchmark coverage for this model, so the overall score is conservative.

## Agentic Benchmarks

| Benchmark | Score | Source |
|-----------|-------|--------|
| [τ²-bench results](/benchmarks/tau2-bench) | 61.1% | [Artificial Analysis model benchmarks](https://artificialanalysis.ai/models/kimi-k2) |

## Coding Benchmarks

| Benchmark | Score | Source |
|-----------|-------|--------|
| [SWE-bench Verified](/benchmarks/swe-bench-verified) | 65.8% | [Kimi-K2-Instruct model card](https://huggingface.co/moonshotai/Kimi-K2-Instruct) |
| [LiveCodeBench v6](/benchmarks/livecodebench-v6) | 53.7% | [Kimi-K2-Instruct model card](https://huggingface.co/moonshotai/Kimi-K2-Instruct) |
| [SWE Multilingual](/benchmarks/swe-bench-multilingual) | 47.3% | [Kimi-K2-Instruct model card](https://huggingface.co/moonshotai/Kimi-K2-Instruct) |

## Multimodal & Grounded Benchmarks

| Benchmark | Score | Source |
|-----------|-------|--------|
| [Design Arena Website](/benchmarks/designarenawebsite) | 1052 | [OpenRouter model benchmarks](https://openrouter.ai/moonshotai/kimi-k2/benchmarks) |

## Reasoning Benchmarks

| Benchmark | Score | Source |
|-----------|-------|--------|
| [AA-LCR](/benchmarks/lcr) | 53.0% | [Artificial Analysis model benchmarks](https://artificialanalysis.ai/models/kimi-k2) |
| [CritPt](/benchmarks/critpt) | 0.0% | [Artificial Analysis model benchmarks](https://artificialanalysis.ai/models/kimi-k2) |

## Knowledge Benchmarks

| Benchmark | Score | Source |
|-----------|-------|--------|
| [MMLU](/benchmarks/mmlu) | 89.5% | [Kimi-K2-Instruct model card](https://huggingface.co/moonshotai/Kimi-K2-Instruct) |
| [GPQA](/benchmarks/gpqa) | 75.1% | [Kimi-K2-Instruct model card](https://huggingface.co/moonshotai/Kimi-K2-Instruct) |
| [SuperGPQA](/benchmarks/supergpqa) | 57.2% | [Kimi-K2-Instruct model card](https://huggingface.co/moonshotai/Kimi-K2-Instruct) |
| [MMLU-Pro](/benchmarks/mmlu-pro) | 81.1% | [Kimi-K2-Instruct model card](https://huggingface.co/moonshotai/Kimi-K2-Instruct) |
| [SimpleQA](/benchmarks/simpleqa) | 31% | [Kimi-K2-Instruct model card](https://huggingface.co/moonshotai/Kimi-K2-Instruct) |
| [HLE](/benchmarks/hle) | 4.7% | [Kimi-K2-Instruct model card](https://huggingface.co/moonshotai/Kimi-K2-Instruct) |
| [Artificial Analysis Intelligence Index](/benchmarks/artificialanalysis) | 12.7% | [Artificial Analysis model benchmarks](https://artificialanalysis.ai/models/kimi-k2) |
| [AA-GPQA Diamond](/benchmarks/gpqa-diamond) | 76.6% | [Artificial Analysis model benchmarks](https://artificialanalysis.ai/models/kimi-k2) |
| [AA-HLE](/benchmarks/aahle) | 7.4% | [Artificial Analysis model benchmarks](https://artificialanalysis.ai/models/kimi-k2) |
| [AA-Omniscience Index](/benchmarks/aaomniscienceindex) | -28.3% | [Artificial Analysis model benchmarks](https://artificialanalysis.ai/models/kimi-k2) |
| [AA-Omniscience Accuracy](/benchmarks/omniscienceaccuracy) | 27.4% | [Artificial Analysis model benchmarks](https://artificialanalysis.ai/models/kimi-k2) |
| [AA-Omniscience Hallucination Rate](/benchmarks/omnisciencehallucinationrate) | 76.6% | [Artificial Analysis model benchmarks](https://artificialanalysis.ai/models/kimi-k2) |

## Instruction Following Benchmarks

| Benchmark | Score | Source |
|-----------|-------|--------|
| [IFEval](/benchmarks/ifeval) | 89.8% | [Kimi-K2-Instruct model card](https://huggingface.co/moonshotai/Kimi-K2-Instruct) |
| [AA-IFBench](/benchmarks/ifbench) | 41.5% | [Artificial Analysis model benchmarks](https://artificialanalysis.ai/models/kimi-k2) |

## Mathematics Benchmarks

| Benchmark | Score | Source |
|-----------|-------|--------|
| [AIME 2024](/benchmarks/aime2024) | 69.6% | [Kimi-K2-Instruct model card](https://huggingface.co/moonshotai/Kimi-K2-Instruct) |
| [AIME 2025](/benchmarks/aime2025) | 49.5% | [Kimi-K2-Instruct model card](https://huggingface.co/moonshotai/Kimi-K2-Instruct) |
| [MATH-500](/benchmarks/math-500) | 97.4% | [Kimi-K2-Instruct model card](https://huggingface.co/moonshotai/Kimi-K2-Instruct) |

## Other Moonshot AI Models

- [Kimi K3](/models/kimi-k3) - Score: 69.87
- [Kimi K2.6](/models/kimi-2-6) - Score: 58.4
- [Kimi K2.5 (Reasoning)](/models/kimi-k2-5-reasoning) - Score: 56.09
- [Kimi K2.5](/models/kimi-k2-5) - Score: 53.05
- [Kimi K2.7 Code](/models/kimi-k2-7-code) - Score: 49.52
- [Moonshot v1](/models/moonshot-v1) - Score: not computed
- [Kimi-Audio 7B](/models/kimi-audio-7b) - Score: not computed