# Gemini 3.5 Flash Lite vs Phi-4

> Gemini 3.5 Flash Lite is the stronger model overall, scoring 41.5 to 31.2 on the Noometry Index. Phi-4 costs 9.7× less per token, which makes it the better buy when Gemini 3.5 Flash Lite's lead doesn't matter for your workload.

- Canonical page: https://noometry.com/compare/gemini-3-5-flash-lite-vs-phi-4
- Last updated: 2026-10-11
- Shared benchmarks: 21

## Summary

- They share 21 benchmarks with published results for both. Gemini 3.5 Flash Lite scores higher in 9 categories and Phi-4 in 0 categories; 9 gaps are clear of the uncertainty.
- The widest gap is in writing & preference, where Gemini 3.5 Flash Lite leads 64.2 to 40.5.
- The biggest single-benchmark swing is OTIS Mock AIME 2024-2025: 71.1% for Gemini 3.5 Flash Lite and 13.8% for Phi-4.
- Phi-4 is cheaper at $0.07 / $0.14 per million input/output tokens, against $0.30 / $2.50 for Gemini 3.5 Flash Lite.
- Gemini 3.5 Flash Lite accepts more context: 1.05M tokens versus 128K.
- Phi-4 has downloadable open weights; the other is API-only.

## Snapshot

| | Gemini 3.5 Flash Lite | Phi-4 |
|---|---|---|
| Provider | Google | Microsoft |
| Noometry Index | 41.5 | 31.2 |
| Rank | 133 | 279 |
| Context | 1.05M | 128K |
| Input $/M | $0.30 | $0.07 |
| Output $/M | $2.50 | $0.14 |
| Weights | Proprietary | Open |

## Coding

- Gemini 3.5 Flash Lite: 42.4 (#104)
- Phi-4: 34.4 (#239)

| Benchmark | Gemini 3.5 Flash Lite | Phi-4 |
|---|---|---|
| LMArena Coding | 1453 | 1231 |
| LMArena WebDev | 1440 | — |
| SciCode | 41.3% | — |
| WeirdML | 39% | — |
| BigCodeBench Instruct | — | 45.5% |
| LiveBench Coding | — | 30.7% |
| BigCodeBench Complete | — | 55.4% |
| ALE-Bench | 765.27 | — |

## Agentic & Tool Use

- Gemini 3.5 Flash Lite: 25.7 (#107)
- Phi-4: 22.8 (#128)

| Benchmark | Gemini 3.5 Flash Lite | Phi-4 |
|---|---|---|
| APEX-Agents | 29.3% | — |
| Berkeley Function Calling Leaderboard | — | 28.8% |
| BALROG | — | 11.6% |
| GDP.pdf | 10% | — |

## Reasoning

- Gemini 3.5 Flash Lite: 27.8 (#114)
- Phi-4: 17.7 (#291)

| Benchmark | Gemini 3.5 Flash Lite | Phi-4 |
|---|---|---|
| Chess Puzzles | 22% | 1% |
| LMArena Hard Prompts | 1441 | 1220 |
| Epoch Capabilities Index | 145.13 | 130.42 |
| ARC-AGI-2 | 10.3% | — |
| NYT Connections (extended) | 60.4% | — |
| ARC-AGI-1 | 53.5% | — |
| CritPt | 0% | — |
| LiveBench Reasoning | — | 47.8% |
| Mystery Game Puzzles | 19% | — |
| DTBench | 83.5% | — |
| LiveBench Data Analysis | — | 45.2% |
| LMCA | 37.6% | — |
| LiveBench | — | 41.6% |

## Math

- Gemini 3.5 Flash Lite: 25.9 (#264)
- Phi-4: 20.8 (#285)

| Benchmark | Gemini 3.5 Flash Lite | Phi-4 |
|---|---|---|
| OTIS Mock AIME 2024-2025 | 71.1% | 13.8% |
| LMArena Math | 1433 | 1246 |
| FrontierMath (Tiers 1-3) | 26% | — |
| FrontierMath Tier 4 | 0% | — |
| ProofBench | 13% | — |
| LiveBench Math | — | 42% |
| MATH Level 5 | — | 64.9% |

## Knowledge

- Gemini 3.5 Flash Lite: 50.1 (#72)
- Phi-4: 32.6 (#209)

| Benchmark | Gemini 3.5 Flash Lite | Phi-4 |
|---|---|---|
| GPQA Diamond | 83.3% | 56.1% |
| LMArena Expert | 1435 | 1203 |
| Confabulations | — | 29.4% |
| Vectara Hallucination Rate | — | 3.7% |
| MMLU | — | 84.8% |

## Multimodal

- Gemini 3.5 Flash Lite: 41.1 (#40)
- Phi-4: —

| Benchmark | Gemini 3.5 Flash Lite | Phi-4 |
|---|---|---|
| LMArena Vision | 1268 | — |

## Multilingual

- Gemini 3.5 Flash Lite: 53.5 (#63)
- Phi-4: 37.2 (#237)

| Benchmark | Gemini 3.5 Flash Lite | Phi-4 |
|---|---|---|
| LMArena Non-English | 1427 | 1197 |
| LMArena Chinese | 1469 | 1212 |
| LMArena French | 1451 | 1224 |
| LMArena German | 1454 | 1222 |
| LMArena Japanese | 1427 | 1158 |
| LMArena Korean | 1407 | 1151 |
| LMArena Russian | 1443 | 1209 |
| LMArena Spanish | 1442 | 1234 |

## Instruction Following

- Gemini 3.5 Flash Lite: 74.8 (#85)
- Phi-4: 60.4 (#251)

| Benchmark | Gemini 3.5 Flash Lite | Phi-4 |
|---|---|---|
| LMArena Instruction Following | 1420 | 1201 |
| LiveBench Instruction Following | — | 58.4% |

## Long Context

- Gemini 3.5 Flash Lite: 43.9 (#84)
- Phi-4: 36.9 (#226)

| Benchmark | Gemini 3.5 Flash Lite | Phi-4 |
|---|---|---|
| LMArena Longer Query | 1435 | 1217 |

## Writing & Preference

- Gemini 3.5 Flash Lite: 64.2 (#57)
- Phi-4: 40.5 (#244)

| Benchmark | Gemini 3.5 Flash Lite | Phi-4 |
|---|---|---|
| LMArena Text | 1435 | 1217 |
| LMArena Creative Writing | 1420 | 1182 |
| LMArena Multi-Turn | 1445 | 1206 |
| Short-Story Creative Writing | — | 62.6% |
| EQ-Bench Creative Writing | 1559 | — |
| LiveBench Language | — | 25.6% |

## FAQ

### Is Gemini 3.5 Flash Lite better than Phi-4?

Gemini 3.5 Flash Lite is the stronger model overall, scoring 41.5 to 31.2 on the Noometry Index. Phi-4 costs 9.7× less per token, which makes it the better buy when Gemini 3.5 Flash Lite's lead doesn't matter for your workload.

### Which is cheaper, Gemini 3.5 Flash Lite or Phi-4?

Phi-4 is cheaper. It lists at $0.07 per million input tokens and $0.14 per million output tokens; Gemini 3.5 Flash Lite lists at $0.30 and $2.50.

### Is Gemini 3.5 Flash Lite or Phi-4 better for coding?

Gemini 3.5 Flash Lite scores higher on coding benchmarks: 42.4 versus 34.4 in the Noometry coding category.

### Which has the bigger context window?

Gemini 3.5 Flash Lite does, with 1.05M tokens against 128K.

### How many benchmarks do Gemini 3.5 Flash Lite and Phi-4 share?

21 benchmarks have published results for both models. Gemini 3.5 Flash Lite has 39 scored results on Noometry and Phi-4 has 37.