# Kimi K2 (Jul 2025) vs o4-mini

> Kimi K2 (Jul 2025) and o4-mini score almost the same on the Noometry Index (40.9 vs 41.6), so choose on price, context window or the category you care about most.

- Canonical page: https://noometry.com/compare/kimi-k2-vs-o4-mini
- Last updated: 2026-10-10
- Shared benchmarks: 39

## Summary

- They share 39 benchmarks with published results for both. Kimi K2 (Jul 2025) scores higher in 3 categories and o4-mini in 6 categories; 7 gaps are clear of the uncertainty.
- The widest gap is in writing & preference, where Kimi K2 (Jul 2025) leads 62.5 to 54.2.
- The biggest single-benchmark swing is SWE-bench Verified (bash only): 63.4% for Kimi K2 (Jul 2025) and 45% for o4-mini.
- Kimi K2 (Jul 2025) is cheaper at $0.57 / $2.30 per million input/output tokens, against $1.10 / $4.40 for o4-mini.
- Kimi K2 (Jul 2025) accepts more context: 262K tokens versus 200K.
- Kimi K2 (Jul 2025) has downloadable open weights; the other is API-only.

## Snapshot

| | Kimi K2 (Jul 2025) | o4-mini |
|---|---|---|
| Provider | Moonshot AI | OpenAI |
| Noometry Index | 40.9 | 41.6 |
| Rank | 134 | 119 |
| Context | 262K | 200K |
| Input $/M | $0.57 | $1.10 |
| Output $/M | $2.30 | $4.40 |
| Weights | Open | Proprietary |

## Coding

- Kimi K2 (Jul 2025): 43.1 (#83)
- o4-mini: 41.6 (#110)

| Benchmark | Kimi K2 (Jul 2025) | o4-mini |
|---|---|---|
| SWE-bench Verified (bash only) | 63.4% | 45% |
| Aider Polyglot | 59.1% | 72% |
| GSO | 4.9% | 3.6% |
| WeirdML | 42.8% | 52.6% |
| LMArena Coding | 1399 | 1369 |
| ALE-Bench | 597.5 | 826.17 |
| CadEval | — | 62% |
| AlgoTune | — | 1.72 |

## Agentic & Tool Use

- Kimi K2 (Jul 2025): 32.2 (#59)
- o4-mini: 32.3 (#57)

| Benchmark | Kimi K2 (Jul 2025) | o4-mini |
|---|---|---|
| Berkeley Function Calling Leaderboard | 59.1% | 53.2% |
| METR Time Horizons | 59.2% | 63.9% |
| Terminal-Bench | 35.7% | — |
| GDPval | — | 25.3% |

## Reasoning

- Kimi K2 (Jul 2025): 22.7 (#173)
- o4-mini: 24.6 (#143)

| Benchmark | Kimi K2 (Jul 2025) | o4-mini |
|---|---|---|
| SimpleBench | 26.3% | 38.7% |
| Kagi LLM Benchmark | 64.4% | 67.6% |
| LMArena Hard Prompts | 1384 | 1351 |
| Epoch Capabilities Index | 146.02 | 145.63 |
| ForecastBench | 60.2 | 61.8 |
| ARC-AGI-2 | — | 6.1% |
| ARC-AGI-1 | — | 58.7% |
| CritPt | — | 0.6% |
| Chess Puzzles | — | 26% |
| EnigmaEval | — | 9.2% |
| Mystery Game Puzzles | — | 5% |
| DTBench | — | 77.6% |
| LMCA | — | 26.5% |

## Math

- Kimi K2 (Jul 2025): 40.1 (#81)
- o4-mini: 40.2 (#79)

| Benchmark | Kimi K2 (Jul 2025) | o4-mini |
|---|---|---|
| Omni-MATH | 65.4% | 72% |
| LMArena Math | 1396 | 1390 |
| FrontierMath (Feb 2025 set) | 21.4% | 24.8% |
| FrontierMath Tier 4 (v1) | 0% | 6.3% |
| FrontierMath (Tiers 1-3) | — | 36.1% |
| FrontierMath Tier 4 | — | 4.9% |
| OTIS Mock AIME 2024-2025 | — | 81.7% |
| MATH Level 5 | — | 97.8% |

## Knowledge

- Kimi K2 (Jul 2025): 36.9 (#153)
- o4-mini: 43.5 (#83)

| Benchmark | Kimi K2 (Jul 2025) | o4-mini |
|---|---|---|
| MMLU-Pro | 81.9% | 82% |
| Confabulations | 20.4% | 15.8% |
| Vectara Hallucination Rate | 17.9% | 18.6% |
| GPQA (HELM) | 65.3% | 73.5% |
| LMArena Expert | 1363 | 1344 |
| GPQA Diamond | — | 79.6% |
| Humanity's Last Exam | — | 18.1% |
| SimpleQA Verified | — | 19.6% |

## Multimodal

- Kimi K2 (Jul 2025): —
- o4-mini: 39.5 (#46)

| Benchmark | Kimi K2 (Jul 2025) | o4-mini |
|---|---|---|
| LMArena Vision | — | 1194 |
| GeoBench | — | 64% |
| VPCT | — | 57.5% |

## Multilingual

- Kimi K2 (Jul 2025): 49.8 (#125)
- o4-mini: 47.2 (#149)

| Benchmark | Kimi K2 (Jul 2025) | o4-mini |
|---|---|---|
| LMArena Non-English | 1372 | 1336 |
| LMArena Chinese | 1416 | 1355 |
| LMArena French | 1379 | 1362 |
| LMArena German | 1386 | 1336 |
| LMArena Japanese | 1349 | 1308 |
| LMArena Korean | 1324 | 1313 |
| LMArena Russian | 1384 | 1334 |
| LMArena Spanish | 1383 | 1346 |

## Instruction Following

- Kimi K2 (Jul 2025): 71.3 (#151)
- o4-mini: 75.4 (#63)

| Benchmark | Kimi K2 (Jul 2025) | o4-mini |
|---|---|---|
| IFEval | 85% | 92.8% |
| LMArena Instruction Following | 1347 | 1321 |

## Long Context

- Kimi K2 (Jul 2025): 41.2 (#140)
- o4-mini: 45.5 (#31)

| Benchmark | Kimi K2 (Jul 2025) | o4-mini |
|---|---|---|
| Fiction.LiveBench | 66.7% | 77.8% |
| LMArena Longer Query | 1353 | 1315 |
| CL-bench | 17.6% | — |

## Writing & Preference

- Kimi K2 (Jul 2025): 62.5 (#76)
- o4-mini: 54.2 (#148)

| Benchmark | Kimi K2 (Jul 2025) | o4-mini |
|---|---|---|
| LMArena Text | 1380 | 1353 |
| LMArena Creative Writing | 1349 | 1294 |
| Short-Story Creative Writing | 85.6% | 75% |
| WildBench | 86.2% | 85.4% |
| LMArena Multi-Turn | 1371 | 1350 |
| EQ-Bench Creative Writing | 1666 | — |

## FAQ

### Is Kimi K2 (Jul 2025) better than o4-mini?

Kimi K2 (Jul 2025) and o4-mini score almost the same on the Noometry Index (40.9 vs 41.6), so choose on price, context window or the category you care about most.

### Which is cheaper, Kimi K2 (Jul 2025) or o4-mini?

Kimi K2 (Jul 2025) is cheaper. It lists at $0.57 per million input tokens and $2.30 per million output tokens; o4-mini lists at $1.10 and $4.40.

### Is Kimi K2 (Jul 2025) or o4-mini better for coding?

Kimi K2 (Jul 2025) scores higher on coding benchmarks: 43.1 versus 41.6 in the Noometry coding category.

### Which has the bigger context window?

Kimi K2 (Jul 2025) does, with 262K tokens against 200K.

### How many benchmarks do Kimi K2 (Jul 2025) and o4-mini share?

39 benchmarks have published results for both models. Kimi K2 (Jul 2025) has 42 scored results on Noometry and o4-mini has 60.