# Deepseek Coder v2 vs Kimi K3

> Kimi K3 is the stronger model overall, scoring 59.5 to 35.3 on the Noometry Index.

- Canonical page: https://noometry.com/compare/deepseek-coder-v2-vs-kimi-k3
- Last updated: 2026-10-10
- Shared benchmarks: 17

## Summary

- They share 17 benchmarks with published results for both. Deepseek Coder v2 scores higher in 0 categories and Kimi K3 in 8 categories; 8 gaps are clear of the uncertainty.
- The widest gap is in math, where Kimi K3 leads 74.0 to 32.3.

## Snapshot

| | Deepseek Coder v2 | Kimi K3 |
|---|---|---|
| Provider | DeepSeek | Moonshot AI |
| Noometry Index | 35.3 | 59.5 |
| Rank | 217 | 12 |
| Context | — | 1.05M |
| Input $/M | — | $3 |
| Output $/M | — | $15 |
| Weights | Open | Open |

## Coding

- Deepseek Coder v2: 37.5 (#178)
- Kimi K3: 60.6 (#8)

| Benchmark | Deepseek Coder v2 | Kimi K3 |
|---|---|---|
| LMArena Coding | 1251 | 1509 |
| DeepSWE | — | 68.5% |
| FrontierCode | — | 44.2% |
| LMArena WebDev | — | 1659 |
| FrontierSWE | — | 25.9% |
| SciCode | — | 59.5% |
| WeirdML | — | 82.6% |
| BigCodeBench Instruct | 48.2% | — |
| BigCodeBench Complete | 59.7% | — |
| ALE-Bench | — | 1,524 |
| HumanEval+ | 82.3% | — |
| MBPP+ | 75.1% | — |

## Agentic & Tool Use

- Deepseek Coder v2: —
- Kimi K3: 41.8 (#19)

| Benchmark | Deepseek Coder v2 | Kimi K3 |
|---|---|---|
| APEX-Agents | — | 50.6% |
| τ²-bench Banking | — | 37.1% |
| PostTrainBench | — | 32% |
| GBAEval | — | 48.3% |
| GDP.pdf | — | 19% |
| Vending-Bench 2 | — | 5,165 |

## Reasoning

- Deepseek Coder v2: 22.6 (#175)
- Kimi K3: 63.0 (#15)

| Benchmark | Deepseek Coder v2 | Kimi K3 |
|---|---|---|
| LMArena Hard Prompts | 1207 | 1495 |
| ARC-AGI-2 | — | 60.4% |
| SimpleBench | — | 60.7% |
| NYT Connections (extended) | — | 93.6% |
| ARC-AGI-1 | — | 94.5% |
| CritPt | — | 23.4% |
| Chess Puzzles | — | 39% |
| Mystery Game Puzzles | — | 26% |
| DTBench | — | 91.2% |
| LMCA | — | 52.7% |
| Surface Evolver Bench | — | 95% |
| Epoch Capabilities Index | — | 157.68 |
| ForecastBench | — | 61.1 |
| WinoGrande | 83.7% | — |

## Math

- Deepseek Coder v2: 32.3 (#188)
- Kimi K3: 74.0 (#13)

| Benchmark | Deepseek Coder v2 | Kimi K3 |
|---|---|---|
| LMArena Math | 1241 | 1493 |
| FrontierMath (Tiers 1-3) | — | 72.2% |
| FrontierMath Tier 4 | — | 39% |
| MathArena Final-Answer Competitions | — | 87.8% |
| OTIS Mock AIME 2024-2025 | — | 97.2% |
| ProofBench | — | 87% |
| GSM8K | 94.5% | — |

## Knowledge

- Deepseek Coder v2: 31.7 (#209)
- Kimi K3: 63.2 (#19)

| Benchmark | Deepseek Coder v2 | Kimi K3 |
|---|---|---|
| LMArena Expert | 1181 | 1520 |
| GPQA Diamond | — | 93.1% |
| SimpleQA Verified | — | 50.6% |
| ARC (AI2) Challenge | 64.3% | — |

## Multimodal

- Deepseek Coder v2: —
- Kimi K3: 38.4 (#58)

| Benchmark | Deepseek Coder v2 | Kimi K3 |
|---|---|---|
| Blueprint-Bench 2 | — | 29.5% |
| Furniture Assembly | — | 34.2% |

## Multilingual

- Deepseek Coder v2: 36.4 (#235)
- Kimi K3: 56.5 (#20)

| Benchmark | Deepseek Coder v2 | Kimi K3 |
|---|---|---|
| LMArena Non-English | 1182 | 1466 |
| LMArena Chinese | 1201 | 1528 |
| LMArena French | 1184 | 1488 |
| LMArena German | 1164 | 1474 |
| LMArena Japanese | 1126 | 1492 |
| LMArena Korean | 1104 | 1465 |
| LMArena Russian | 1188 | 1481 |
| LMArena Spanish | 1153 | 1478 |

## Instruction Following

- Deepseek Coder v2: 61.9 (#237)
- Kimi K3: 77.8 (#13)

| Benchmark | Deepseek Coder v2 | Kimi K3 |
|---|---|---|
| LMArena Instruction Following | 1180 | 1481 |

## Long Context

- Deepseek Coder v2: 37.0 (#219)
- Kimi K3: 45.8 (#27)

| Benchmark | Deepseek Coder v2 | Kimi K3 |
|---|---|---|
| LMArena Longer Query | 1219 | 1494 |

## Writing & Preference

- Deepseek Coder v2: 38.4 (#249)
- Kimi K3: 76.7 (#4)

| Benchmark | Deepseek Coder v2 | Kimi K3 |
|---|---|---|
| LMArena Text | 1191 | 1476 |
| LMArena Creative Writing | 1120 | 1452 |
| LMArena Multi-Turn | 1176 | 1488 |
| EQ-Bench Creative Writing | — | 2082 |
| EQ-Bench 4 | — | 1339 |

## FAQ

### Is Deepseek Coder v2 better than Kimi K3?

Kimi K3 is the stronger model overall, scoring 59.5 to 35.3 on the Noometry Index.

### Is Deepseek Coder v2 or Kimi K3 better for coding?

Kimi K3 scores higher on coding benchmarks: 60.6 versus 37.5 in the Noometry coding category.

### How many benchmarks do Deepseek Coder v2 and Kimi K3 share?

17 benchmarks have published results for both models. Deepseek Coder v2 has 24 scored results on Noometry and Kimi K3 has 53.