# GPT-4 Turbo vs Phi-4 Mini

> GPT-4 Turbo and Phi-4 Mini score almost the same on the Noometry Index (30.3 vs 30.4), so choose on price, context window or the category you care about most.

- Canonical page: https://noometry.com/compare/gpt-4-turbo-vs-phi-4-mini
- Last updated: 2026-10-10
- Shared benchmarks: 0

## Summary

- The widest gap is in reasoning, where Phi-4 Mini leads 21.7 to 14.9.
- Phi-4 Mini is cheaper at $0.075 / $0.30 per million input/output tokens, against $10 / $30 for GPT-4 Turbo.
- Phi-4 Mini has downloadable open weights; the other is API-only.

## Snapshot

| | GPT-4 Turbo | Phi-4 Mini |
|---|---|---|
| Provider | OpenAI | Microsoft |
| Noometry Index | 30.3 | 30.4 |
| Rank | 281 | 280 |
| Context | 128K | 128K |
| Input $/M | $10 | $0.075 |
| Output $/M | $30 | $0.30 |
| Weights | Proprietary | Open |

## Coding

- GPT-4 Turbo: 33.4 (#247)
- Phi-4 Mini: 27.6 (#312)

| Benchmark | GPT-4 Turbo | Phi-4 Mini |
|---|---|---|
| SciCode | — | 10.8% |
| WeirdML | 18% | — |
| BigCodeBench Instruct | 48.2% | — |
| LMArena Coding | 1268 | — |
| BigCodeBench Complete | 58.2% | — |
| HumanEval+ | 86.6% | — |
| MBPP+ | 73.3% | — |

## Agentic & Tool Use

- GPT-4 Turbo: —
- Phi-4 Mini: —

| Benchmark | GPT-4 Turbo | Phi-4 Mini |
|---|---|---|
| METR Time Horizons | 36.7% | — |

## Reasoning

- GPT-4 Turbo: 14.9 (#311)
- Phi-4 Mini: 21.7 (#190)

| Benchmark | GPT-4 Turbo | Phi-4 Mini |
|---|---|---|
| SimpleBench | 25.1% | — |
| CritPt | — | 0% |
| Chess Puzzles | 6% | — |
| LMArena Hard Prompts | 1251 | — |
| DTBench | 61.6% | — |
| LMCA | 9.8% | — |
| Epoch Capabilities Index | 127.25 | — |
| ForecastBench | 59.4 | — |

## Math

- GPT-4 Turbo: 8.6 (#316)
- Phi-4 Mini: —

| Benchmark | GPT-4 Turbo | Phi-4 Mini |
|---|---|---|
| FrontierMath (Tiers 1-3) | 0.7% | — |
| OTIS Mock AIME 2024-2025 | 6.7% | — |
| LMArena Math | 1272 | — |
| MATH Level 5 | 46.7% | — |

## Knowledge

- GPT-4 Turbo: 24.3 (#263)
- Phi-4 Mini: 24.8 (#261)

| Benchmark | GPT-4 Turbo | Phi-4 Mini |
|---|---|---|
| GPQA Diamond | 46.6% | — |
| Confabulations | 28.4% | — |
| Vectara Hallucination Rate | — | 23.5% |
| LMArena Expert | 1223 | — |
| MMLU | 81.3% | — |

## Multimodal

- GPT-4 Turbo: 30.0 (#104)
- Phi-4 Mini: —

| Benchmark | GPT-4 Turbo | Phi-4 Mini |
|---|---|---|
| LMArena Vision | 1090 | — |

## Multilingual

- GPT-4 Turbo: 40.7 (#211)
- Phi-4 Mini: —

| Benchmark | GPT-4 Turbo | Phi-4 Mini |
|---|---|---|
| LMArena Non-English | 1245 | — |
| LMArena Chinese | 1242 | — |
| LMArena French | 1275 | — |
| LMArena German | 1259 | — |
| LMArena Japanese | 1194 | — |
| LMArena Korean | 1187 | — |
| LMArena Russian | 1259 | — |
| LMArena Spanish | 1260 | — |

## Instruction Following

- GPT-4 Turbo: 66.0 (#211)
- Phi-4 Mini: —

| Benchmark | GPT-4 Turbo | Phi-4 Mini |
|---|---|---|
| LMArena Instruction Following | 1250 | — |

## Long Context

- GPT-4 Turbo: 38.0 (#200)
- Phi-4 Mini: —

| Benchmark | GPT-4 Turbo | Phi-4 Mini |
|---|---|---|
| LMArena Longer Query | 1254 | — |

## Writing & Preference

- GPT-4 Turbo: 47.9 (#201)
- Phi-4 Mini: —

| Benchmark | GPT-4 Turbo | Phi-4 Mini |
|---|---|---|
| LMArena Text | 1272 | — |
| LMArena Creative Writing | 1269 | — |
| LMArena Multi-Turn | 1267 | — |

## FAQ

### Is GPT-4 Turbo better than Phi-4 Mini?

GPT-4 Turbo and Phi-4 Mini score almost the same on the Noometry Index (30.3 vs 30.4), so choose on price, context window or the category you care about most.

### Which is cheaper, GPT-4 Turbo or Phi-4 Mini?

Phi-4 Mini is cheaper. It lists at $0.075 per million input tokens and $0.30 per million output tokens; GPT-4 Turbo lists at $10 and $30.

### Is GPT-4 Turbo or Phi-4 Mini better for coding?

GPT-4 Turbo scores higher on coding benchmarks: 33.4 versus 27.6 in the Noometry coding category.

### Which has the bigger context window?

Both accept 128K tokens.

### How many benchmarks do GPT-4 Turbo and Phi-4 Mini share?

0 benchmarks have published results for both models. GPT-4 Turbo has 36 scored results on Noometry and Phi-4 Mini has 3.