# Grok-3 mini vs Phi-4

> Grok-3 mini is the stronger model overall, scoring 41.2 to 31.2 on the Noometry Index.

- Canonical page: https://noometry.com/compare/grok-3-mini-vs-phi-4
- Last updated: 2026-10-11
- Shared benchmarks: 23

## Summary

- They share 23 benchmarks with published results for both. Grok-3 mini scores higher in 7 categories and Phi-4 in 1 category; 8 gaps are clear of the uncertainty.
- The widest gap is in math, where Grok-3 mini leads 42.1 to 20.8.
- The biggest single-benchmark swing is OTIS Mock AIME 2024-2025: 77.8% for Grok-3 mini and 13.8% for Phi-4.
- Phi-4 has downloadable open weights; the other is API-only.

## Snapshot

| | Grok-3 mini | Phi-4 |
|---|---|---|
| Provider | xAI | Microsoft |
| Noometry Index | 41.2 | 31.2 |
| Rank | 141 | 279 |
| Context | — | 128K |
| Input $/M | — | $0.07 |
| Output $/M | — | $0.14 |
| Weights | Proprietary | Open |

## Coding

- Grok-3 mini: 40.8 (#131)
- Phi-4: 34.4 (#239)

| Benchmark | Grok-3 mini | Phi-4 |
|---|---|---|
| LMArena Coding | 1379 | 1231 |
| Aider Polyglot | 49.3% | — |
| WeirdML | 42.6% | — |
| BigCodeBench Instruct | — | 45.5% |
| LiveBench Coding | — | 30.7% |
| BigCodeBench Complete | — | 55.4% |

## Agentic & Tool Use

- Grok-3 mini: —
- Phi-4: 22.8 (#128)

| Benchmark | Grok-3 mini | Phi-4 |
|---|---|---|
| Berkeley Function Calling Leaderboard | — | 28.8% |
| BALROG | — | 11.6% |

## Reasoning

- Grok-3 mini: 13.6 (#334)
- Phi-4: 17.7 (#291)

| Benchmark | Grok-3 mini | Phi-4 |
|---|---|---|
| LMArena Hard Prompts | 1375 | 1220 |
| Epoch Capabilities Index | 140.35 | 130.42 |
| ARC-AGI-2 | 0.4% | — |
| Kagi LLM Benchmark | 61.3% | — |
| ARC-AGI-1 | 16.5% | — |
| Chess Puzzles | — | 1% |
| LiveBench Reasoning | — | 47.8% |
| LiveBench Data Analysis | — | 45.2% |
| LiveBench | — | 41.6% |

## Math

- Grok-3 mini: 42.1 (#85)
- Phi-4: 20.8 (#285)

| Benchmark | Grok-3 mini | Phi-4 |
|---|---|---|
| OTIS Mock AIME 2024-2025 | 77.8% | 13.8% |
| LMArena Math | 1386 | 1246 |
| MATH Level 5 | 90.9% | 64.9% |
| Omni-MATH | 31.8% | — |
| LiveBench Math | — | 42% |
| FrontierMath (Feb 2025 set) | 5.9% | — |

## Knowledge

- Grok-3 mini: 46.4 (#81)
- Phi-4: 32.6 (#209)

| Benchmark | Grok-3 mini | Phi-4 |
|---|---|---|
| GPQA Diamond | 76.3% | 56.1% |
| Confabulations | 10.8% | 29.4% |
| LMArena Expert | 1395 | 1203 |
| MMLU-Pro | 79.9% | — |
| Vectara Hallucination Rate | — | 3.7% |
| GPQA (HELM) | 67.5% | — |
| MMLU | — | 84.8% |

## Multilingual

- Grok-3 mini: 48.1 (#145)
- Phi-4: 37.2 (#237)

| Benchmark | Grok-3 mini | Phi-4 |
|---|---|---|
| LMArena Non-English | 1352 | 1197 |
| LMArena Chinese | 1387 | 1212 |
| LMArena French | 1357 | 1224 |
| LMArena German | 1349 | 1222 |
| LMArena Japanese | 1342 | 1158 |
| LMArena Korean | 1335 | 1151 |
| LMArena Russian | 1353 | 1209 |
| LMArena Spanish | 1381 | 1234 |

## Instruction Following

- Grok-3 mini: 78.5 (#9)
- Phi-4: 60.4 (#251)

| Benchmark | Grok-3 mini | Phi-4 |
|---|---|---|
| LMArena Instruction Following | 1357 | 1201 |
| LiveBench Instruction Following | — | 58.4% |
| IFEval | 95.1% | — |

## Long Context

- Grok-3 mini: 41.0 (#147)
- Phi-4: 36.9 (#226)

| Benchmark | Grok-3 mini | Phi-4 |
|---|---|---|
| LMArena Longer Query | 1372 | 1217 |
| Fiction.LiveBench | 66.7% | — |

## Writing & Preference

- Grok-3 mini: 52.5 (#169)
- Phi-4: 40.5 (#244)

| Benchmark | Grok-3 mini | Phi-4 |
|---|---|---|
| LMArena Text | 1370 | 1217 |
| LMArena Creative Writing | 1342 | 1182 |
| Short-Story Creative Writing | 73.5% | 62.6% |
| LMArena Multi-Turn | 1355 | 1206 |
| WildBench | 65.1% | — |
| LiveBench Language | — | 25.6% |

## FAQ

### Is Grok-3 mini better than Phi-4?

Grok-3 mini is the stronger model overall, scoring 41.2 to 31.2 on the Noometry Index.

### Is Grok-3 mini or Phi-4 better for coding?

Grok-3 mini scores higher on coding benchmarks: 40.8 versus 34.4 in the Noometry coding category.

### How many benchmarks do Grok-3 mini and Phi-4 share?

23 benchmarks have published results for both models. Grok-3 mini has 35 scored results on Noometry and Phi-4 has 37.