# Deepseek Coder v2 vs Phi 3 Small 8k Instruct

> Deepseek Coder v2 is the stronger model overall, scoring 35.3 to 28.8 on the Noometry Index.

- Canonical page: https://noometry.com/compare/deepseek-coder-v2-vs-phi-3-small-8k-instruct
- Last updated: 2026-10-10
- Shared benchmarks: 19

## Summary

- They share 19 benchmarks with published results for both. Deepseek Coder v2 scores higher in 8 categories and Phi 3 Small 8k Instruct in 0 categories; 8 gaps are clear of the uncertainty.
- The widest gap is in coding, where Deepseek Coder v2 leads 37.5 to 27.4.

## Snapshot

| | Deepseek Coder v2 | Phi 3 Small 8k Instruct |
|---|---|---|
| Provider | DeepSeek | Microsoft |
| Noometry Index | 35.3 | 28.8 |
| Rank | 217 | 310 |
| Context | — | — |
| Input $/M | — | — |
| Output $/M | — | — |
| Weights | Open | Open |

## Coding

- Deepseek Coder v2: 37.5 (#178)
- Phi 3 Small 8k Instruct: 27.4 (#313)

| Benchmark | Deepseek Coder v2 | Phi 3 Small 8k Instruct |
|---|---|---|
| LMArena Coding | 1251 | 1102 |
| BigCodeBench Instruct | 48.2% | — |
| LiveBench Coding | — | 20.3% |
| BigCodeBench Complete | 59.7% | — |
| HumanEval+ | 82.3% | — |
| MBPP+ | 75.1% | — |

## Reasoning

- Deepseek Coder v2: 22.6 (#175)
- Phi 3 Small 8k Instruct: 14.2 (#319)

| Benchmark | Deepseek Coder v2 | Phi 3 Small 8k Instruct |
|---|---|---|
| LMArena Hard Prompts | 1207 | 1100 |
| WinoGrande | 83.7% | 81.5% |
| LiveBench Reasoning | — | 15.9% |
| LiveBench Data Analysis | — | 30.3% |
| Adversarial NLI | — | 58.1% |
| BIG-Bench Hard | — | 79.1% |
| HellaSwag | — | 77% |
| LiveBench | — | 24% |

## Math

- Deepseek Coder v2: 32.3 (#188)
- Phi 3 Small 8k Instruct: 25.4 (#247)

| Benchmark | Deepseek Coder v2 | Phi 3 Small 8k Instruct |
|---|---|---|
| LMArena Math | 1241 | 1151 |
| LiveBench Math | — | 17.6% |
| GSM8K | 94.5% | — |

## Knowledge

- Deepseek Coder v2: 31.7 (#209)
- Phi 3 Small 8k Instruct: 28.5 (#237)

| Benchmark | Deepseek Coder v2 | Phi 3 Small 8k Instruct |
|---|---|---|
| LMArena Expert | 1181 | 1067 |
| ARC (AI2) Challenge | 64.3% | 90.7% |
| MMLU | — | 75.7% |
| OpenBookQA | — | 88% |
| TriviaQA | — | 58.1% |

## Multilingual

- Deepseek Coder v2: 36.4 (#235)
- Phi 3 Small 8k Instruct: 28.6 (#267)

| Benchmark | Deepseek Coder v2 | Phi 3 Small 8k Instruct |
|---|---|---|
| LMArena Non-English | 1182 | 1058 |
| LMArena Chinese | 1201 | 1061 |
| LMArena French | 1184 | 1134 |
| LMArena German | 1164 | 1080 |
| LMArena Japanese | 1126 | 966 |
| LMArena Korean | 1104 | 894 |
| LMArena Russian | 1188 | 1111 |
| LMArena Spanish | 1153 | 1111 |

## Instruction Following

- Deepseek Coder v2: 61.9 (#237)
- Phi 3 Small 8k Instruct: 52.2 (#287)

| Benchmark | Deepseek Coder v2 | Phi 3 Small 8k Instruct |
|---|---|---|
| LMArena Instruction Following | 1180 | 1087 |
| LiveBench Instruction Following | — | 47.2% |

## Long Context

- Deepseek Coder v2: 37.0 (#219)
- Phi 3 Small 8k Instruct: 33.0 (#262)

| Benchmark | Deepseek Coder v2 | Phi 3 Small 8k Instruct |
|---|---|---|
| LMArena Longer Query | 1219 | 1088 |

## Writing & Preference

- Deepseek Coder v2: 38.4 (#249)
- Phi 3 Small 8k Instruct: 31.3 (#279)

| Benchmark | Deepseek Coder v2 | Phi 3 Small 8k Instruct |
|---|---|---|
| LMArena Text | 1191 | 1110 |
| LMArena Creative Writing | 1120 | 1083 |
| LMArena Multi-Turn | 1176 | 1068 |
| LiveBench Language | — | 12.9% |

## FAQ

### Is Deepseek Coder v2 better than Phi 3 Small 8k Instruct?

Deepseek Coder v2 is the stronger model overall, scoring 35.3 to 28.8 on the Noometry Index.

### Is Deepseek Coder v2 or Phi 3 Small 8k Instruct better for coding?

Deepseek Coder v2 scores higher on coding benchmarks: 37.5 versus 27.4 in the Noometry coding category.

### How many benchmarks do Deepseek Coder v2 and Phi 3 Small 8k Instruct share?

19 benchmarks have published results for both models. Deepseek Coder v2 has 24 scored results on Noometry and Phi 3 Small 8k Instruct has 32.