# MiniMax-M3 vs Phi-4

> MiniMax-M3 is the stronger model overall, scoring 43.5 to 30.9 on the Noometry Index. Phi-4 costs 6.0× less per token, which makes it the better buy when MiniMax-M3's lead doesn't matter for your workload.

- Canonical page: https://noometry.com/compare/minimax-m3-vs-phi-4
- Last updated: 2026-10-10
- Shared benchmarks: 21

## Summary

- They share 21 benchmarks with published results for both. MiniMax-M3 scores higher in 8 categories and Phi-4 in 1 category; 8 gaps are clear of the uncertainty.
- The widest gap is in knowledge, where MiniMax-M3 leads 58.4 to 32.5.
- The biggest single-benchmark swing is OTIS Mock AIME 2024-2025: 71.1% for MiniMax-M3 and 13.8% for Phi-4.
- Phi-4 is cheaper at $0.07 / $0.14 per million input/output tokens, against $0.30 / $1.20 for MiniMax-M3.
- MiniMax-M3 accepts more context: 1M tokens versus 128K.

## Snapshot

| | MiniMax-M3 | Phi-4 |
|---|---|---|
| Provider | MiniMax | Microsoft |
| Noometry Index | 43.5 | 30.9 |
| Rank | 81 | 269 |
| Context | 1M | 128K |
| Input $/M | $0.30 | $0.07 |
| Output $/M | $1.20 | $0.14 |
| Weights | Open | Open |

## Coding

- MiniMax-M3: 41.5 (#111)
- Phi-4: 33.8 (#237)

| Benchmark | MiniMax-M3 | Phi-4 |
|---|---|---|
| LMArena Coding | 1470 | 1231 |
| FrontierCode | 14.7% | — |
| LMArena WebDev | 1482 | — |
| SciCode | 47.1% | — |
| BigCodeBench Instruct | — | 45.5% |
| LiveBench Coding | — | 30.7% |
| BigCodeBench Complete | — | 55.4% |
| ALE-Bench | 640.02 | — |

## Agentic & Tool Use

- MiniMax-M3: 22.5 (#121)
- Phi-4: 22.6 (#120)

| Benchmark | MiniMax-M3 | Phi-4 |
|---|---|---|
| APEX-Agents | 37.7% | — |
| Berkeley Function Calling Leaderboard | — | 28.8% |
| OSWorld 2.0 | 4.6% | — |
| BALROG | — | 11.6% |
| GBAEval | 0.9% | — |
| Vending-Bench 2 | 2,158 | — |

## Reasoning

- MiniMax-M3: 29.3 (#83)
- Phi-4: 17.0 (#287)

| Benchmark | MiniMax-M3 | Phi-4 |
|---|---|---|
| Chess Puzzles | 14% | 1% |
| LMArena Hard Prompts | 1447 | 1220 |
| Epoch Capabilities Index | 147 | 130.43 |
| SimpleBench | 45.8% | — |
| NYT Connections (extended) | 65.1% | — |
| CritPt | 3.7% | — |
| LiveBench Reasoning | — | 47.8% |
| Mystery Game Puzzles | 8% | — |
| DTBench | 78.9% | — |
| LiveBench Data Analysis | — | 45.2% |
| LMCA | 33.7% | — |
| Surface Evolver Bench | 55% | — |
| ForecastBench | 61.4 | — |
| LiveBench | — | 41.6% |

## Math

- MiniMax-M3: 39.1 (#83)
- Phi-4: 19.6 (#279)

| Benchmark | MiniMax-M3 | Phi-4 |
|---|---|---|
| OTIS Mock AIME 2024-2025 | 71.1% | 13.8% |
| LMArena Math | 1432 | 1246 |
| ProofBench | 18% | — |
| LiveBench Math | — | 42% |
| MATH Level 5 | — | 64.9% |

## Knowledge

- MiniMax-M3: 58.4 (#32)
- Phi-4: 32.5 (#204)

| Benchmark | MiniMax-M3 | Phi-4 |
|---|---|---|
| GPQA Diamond | 90.9% | 56.1% |
| LMArena Expert | 1462 | 1203 |
| Confabulations | — | 29.4% |
| Vectara Hallucination Rate | — | 3.7% |
| MMLU | — | 84.8% |

## Multimodal

- MiniMax-M3: 39.5 (#45)
- Phi-4: —

| Benchmark | MiniMax-M3 | Phi-4 |
|---|---|---|
| LMArena Vision | 1254 | — |
| LMArena Document | 1435 | — |

## Multilingual

- MiniMax-M3: 53.2 (#70)
- Phi-4: 37.4 (#232)

| Benchmark | MiniMax-M3 | Phi-4 |
|---|---|---|
| LMArena Non-English | 1420 | 1197 |
| LMArena Chinese | 1465 | 1211 |
| LMArena French | 1445 | 1223 |
| LMArena German | 1424 | 1222 |
| LMArena Japanese | 1381 | 1158 |
| LMArena Korean | 1372 | 1151 |
| LMArena Russian | 1430 | 1209 |
| LMArena Spanish | 1431 | 1234 |

## Instruction Following

- MiniMax-M3: 75.7 (#56)
- Phi-4: 60.6 (#246)

| Benchmark | MiniMax-M3 | Phi-4 |
|---|---|---|
| LMArena Instruction Following | 1433 | 1201 |
| LiveBench Instruction Following | — | 58.4% |

## Long Context

- MiniMax-M3: 44.2 (#68)
- Phi-4: 36.9 (#221)

| Benchmark | MiniMax-M3 | Phi-4 |
|---|---|---|
| LMArena Longer Query | 1445 | 1217 |

## Writing & Preference

- MiniMax-M3: 62.3 (#79)
- Phi-4: 40.7 (#239)

| Benchmark | MiniMax-M3 | Phi-4 |
|---|---|---|
| LMArena Text | 1433 | 1217 |
| LMArena Creative Writing | 1405 | 1182 |
| LMArena Multi-Turn | 1441 | 1206 |
| Short-Story Creative Writing | — | 62.6% |
| EQ-Bench 4 | 1150 | — |
| LiveBench Language | — | 25.6% |

## FAQ

### Is MiniMax-M3 better than Phi-4?

MiniMax-M3 is the stronger model overall, scoring 43.5 to 30.9 on the Noometry Index. Phi-4 costs 6.0× less per token, which makes it the better buy when MiniMax-M3's lead doesn't matter for your workload.

### Which is cheaper, MiniMax-M3 or Phi-4?

Phi-4 is cheaper. It lists at $0.07 per million input tokens and $0.14 per million output tokens; MiniMax-M3 lists at $0.30 and $1.20.

### Is MiniMax-M3 or Phi-4 better for coding?

MiniMax-M3 scores higher on coding benchmarks: 41.5 versus 33.8 in the Noometry coding category.

### Which has the bigger context window?

MiniMax-M3 does, with 1M tokens against 128K.

### How many benchmarks do MiniMax-M3 and Phi-4 share?

21 benchmarks have published results for both models. MiniMax-M3 has 41 scored results on Noometry and Phi-4 has 37.