# Grok-3 mini vs Kimi K2 Thinking Turbo

> Kimi K2 Thinking Turbo is the stronger model overall, scoring 45.4 to 41.0 on the Noometry Index.

- Canonical page: https://noometry.com/compare/grok-3-mini-vs-kimi-k2-thinking-turbo
- Last updated: 2026-10-10
- Shared benchmarks: 19

## Summary

- They share 19 benchmarks with published results for both. Grok-3 mini scores higher in 2 categories and Kimi K2 Thinking Turbo in 6 categories; 8 gaps are clear of the uncertainty.
- The widest gap is in reasoning, where Kimi K2 Thinking Turbo leads 31.4 to 13.8.
- The biggest single-benchmark swing is GPQA Diamond: 76.3% for Grok-3 mini and 84.2% for Kimi K2 Thinking Turbo.
- Kimi K2 Thinking Turbo has downloadable open weights; the other is API-only.

## Snapshot

| | Grok-3 mini | Kimi K2 Thinking Turbo |
|---|---|---|
| Provider | xAI | Moonshot AI |
| Noometry Index | 41.0 | 45.4 |
| Rank | 130 | 63 |
| Context | — | — |
| Input $/M | — | — |
| Output $/M | — | — |
| Weights | Proprietary | Open |

## Coding

- Grok-3 mini: 40.4 (#126)
- Kimi K2 Thinking Turbo: 38.1 (#166)

| Benchmark | Grok-3 mini | Kimi K2 Thinking Turbo |
|---|---|---|
| LMArena Coding | 1380 | 1454 |
| Aider Polyglot | 49.3% | — |
| LMArena WebDev | — | 1322 |
| WeirdML | 42.6% | — |

## Reasoning

- Grok-3 mini: 13.8 (#323)
- Kimi K2 Thinking Turbo: 31.4 (#72)

| Benchmark | Grok-3 mini | Kimi K2 Thinking Turbo |
|---|---|---|
| LMArena Hard Prompts | 1374 | 1428 |
| ARC-AGI-2 | 0.4% | — |
| Kagi LLM Benchmark | 61.3% | — |
| ARC-AGI-1 | 16.5% | — |
| Chess Puzzles | — | 20% |
| Epoch Capabilities Index | 140.35 | — |

## Math

- Grok-3 mini: 40.3 (#78)
- Kimi K2 Thinking Turbo: 45.5 (#62)

| Benchmark | Grok-3 mini | Kimi K2 Thinking Turbo |
|---|---|---|
| OTIS Mock AIME 2024-2025 | 77.8% | 83.1% |
| LMArena Math | 1386 | 1429 |
| Omni-MATH | 31.8% | — |
| MATH Level 5 | 90.9% | — |
| FrontierMath (Feb 2025 set) | 5.9% | — |

## Knowledge

- Grok-3 mini: 46.3 (#75)
- Kimi K2 Thinking Turbo: 50.9 (#64)

| Benchmark | Grok-3 mini | Kimi K2 Thinking Turbo |
|---|---|---|
| GPQA Diamond | 76.3% | 84.2% |
| LMArena Expert | 1395 | 1439 |
| MMLU-Pro | 79.9% | — |
| Confabulations | 10.8% | — |
| GPQA (HELM) | 67.5% | — |

## Multilingual

- Grok-3 mini: 48.3 (#141)
- Kimi K2 Thinking Turbo: 51.6 (#104)

| Benchmark | Grok-3 mini | Kimi K2 Thinking Turbo |
|---|---|---|
| LMArena Non-English | 1352 | 1398 |
| LMArena Chinese | 1387 | 1457 |
| LMArena French | 1357 | 1425 |
| LMArena German | 1350 | 1390 |
| LMArena Japanese | 1343 | 1356 |
| LMArena Korean | 1336 | 1329 |
| LMArena Russian | 1352 | 1390 |
| LMArena Spanish | 1379 | 1407 |

## Instruction Following

- Grok-3 mini: 78.7 (#8)
- Kimi K2 Thinking Turbo: 74.2 (#103)

| Benchmark | Grok-3 mini | Kimi K2 Thinking Turbo |
|---|---|---|
| LMArena Instruction Following | 1356 | 1403 |
| IFEval | 95.1% | — |

## Long Context

- Grok-3 mini: 41.0 (#142)
- Kimi K2 Thinking Turbo: 43.2 (#99)

| Benchmark | Grok-3 mini | Kimi K2 Thinking Turbo |
|---|---|---|
| LMArena Longer Query | 1372 | 1415 |
| Fiction.LiveBench | 66.7% | — |

## Writing & Preference

- Grok-3 mini: 52.7 (#163)
- Kimi K2 Thinking Turbo: 60.2 (#99)

| Benchmark | Grok-3 mini | Kimi K2 Thinking Turbo |
|---|---|---|
| LMArena Text | 1370 | 1415 |
| LMArena Creative Writing | 1342 | 1373 |
| LMArena Multi-Turn | 1355 | 1414 |
| Short-Story Creative Writing | 73.5% | — |
| WildBench | 65.1% | — |

## FAQ

### Is Grok-3 mini better than Kimi K2 Thinking Turbo?

Kimi K2 Thinking Turbo is the stronger model overall, scoring 45.4 to 41.0 on the Noometry Index.

### Is Grok-3 mini or Kimi K2 Thinking Turbo better for coding?

Grok-3 mini scores higher on coding benchmarks: 40.4 versus 38.1 in the Noometry coding category.

### How many benchmarks do Grok-3 mini and Kimi K2 Thinking Turbo share?

19 benchmarks have published results for both models. Grok-3 mini has 35 scored results on Noometry and Kimi K2 Thinking Turbo has 21.