# Gemini 3.1 Flash Lite vs GPT-4

> Gemini 3.1 Flash Lite is the stronger model overall, scoring 40.4 to 28.9 on the Noometry Index.

- Canonical page: https://noometry.com/compare/gemini-3-1-flash-lite-vs-gpt-4
- Last updated: 2026-10-10
- Shared benchmarks: 25

## Summary

- They share 25 benchmarks with published results for both. Gemini 3.1 Flash Lite scores higher in 8 categories and GPT-4 in 0 categories; 8 gaps are clear of the uncertainty.
- The widest gap is in math, where Gemini 3.1 Flash Lite leads 39.5 to 10.1.
- The biggest single-benchmark swing is OTIS Mock AIME 2024-2025: 80% for Gemini 3.1 Flash Lite and 1.1% for GPT-4.
- Gemini 3.1 Flash Lite is cheaper at $0.25 / $1.50 per million input/output tokens, against $30 / $60 for GPT-4.
- Gemini 3.1 Flash Lite accepts more context: 1.05M tokens versus 8K.

## Snapshot

| | Gemini 3.1 Flash Lite | GPT-4 |
|---|---|---|
| Provider | Google | OpenAI |
| Noometry Index | 40.4 | 28.9 |
| Rank | 141 | 308 |
| Context | 1.05M | 8K |
| Input $/M | $0.25 | $30 |
| Output $/M | $1.50 | $60 |
| Weights | Proprietary | Proprietary |

## Coding

- Gemini 3.1 Flash Lite: 37.5 (#179)
- GPT-4: 31.2 (#277)

| Benchmark | Gemini 3.1 Flash Lite | GPT-4 |
|---|---|---|
| WeirdML | 52.2% | 12.4% |
| LMArena Coding | 1400 | 1254 |
| LMArena WebDev | 1256 | — |
| SciCode | 41.9% | — |
| BigCodeBench Instruct | — | 46% |
| BigCodeBench Complete | — | 57.2% |
| ALE-Bench | 797.73 | — |
| HumanEval+ | — | 79.3% |

## Agentic & Tool Use

- Gemini 3.1 Flash Lite: 29.9 (#72)
- GPT-4: —

| Benchmark | Gemini 3.1 Flash Lite | GPT-4 |
|---|---|---|
| DeepResearch Bench | 37.3% | — |
| METR Time Horizons | — | 36.1% |

## Reasoning

- Gemini 3.1 Flash Lite: 22.3 (#181)
- GPT-4: 17.3 (#285)

| Benchmark | Gemini 3.1 Flash Lite | GPT-4 |
|---|---|---|
| Chess Puzzles | 25% | 4% |
| LMArena Hard Prompts | 1407 | 1241 |
| DTBench | 76.8% | 62.7% |
| LMCA | 35% | 17.1% |
| Epoch Capabilities Index | 144.45 | 125.88 |
| ForecastBench | 54.4 | 57.8 |
| Kagi LLM Benchmark | 67.2% | — |
| NYT Connections (extended) | 8.2% | — |
| CritPt | 1.1% | — |
| EnigmaEval | 3% | — |
| Thematic Generalization | 63.3% | — |
| Mystery Game Puzzles | — | 12% |
| BIG-Bench Hard | — | 75.1% |
| HellaSwag | — | 95.3% |
| WinoGrande | — | 87.5% |

## Math

- Gemini 3.1 Flash Lite: 39.5 (#82)
- GPT-4: 10.1 (#304)

| Benchmark | Gemini 3.1 Flash Lite | GPT-4 |
|---|---|---|
| OTIS Mock AIME 2024-2025 | 80% | 1.1% |
| LMArena Math | 1427 | 1269 |
| FrontierMath (Tiers 1-3) | 27.7% | — |
| MATH Level 5 | — | 23% |
| GSM8K | — | 92% |

## Knowledge

- Gemini 3.1 Flash Lite: 41.7 (#95)
- GPT-4: 18.4 (#277)

| Benchmark | Gemini 3.1 Flash Lite | GPT-4 |
|---|---|---|
| GPQA Diamond | 81.8% | 35.7% |
| LMArena Expert | 1397 | 1211 |
| Humanity's Last Exam | 8.6% | — |
| Vectara Hallucination Rate | 8.2% | — |
| MMLU | — | 86.4% |
| TriviaQA | — | 84.8% |

## Multimodal

- Gemini 3.1 Flash Lite: 38.7 (#57)
- GPT-4: —

| Benchmark | Gemini 3.1 Flash Lite | GPT-4 |
|---|---|---|
| LMArena Vision | 1240 | — |

## Multilingual

- Gemini 3.1 Flash Lite: 52.5 (#82)
- GPT-4: 40.8 (#210)

| Benchmark | Gemini 3.1 Flash Lite | GPT-4 |
|---|---|---|
| LMArena Non-English | 1410 | 1246 |
| LMArena Chinese | 1461 | 1242 |
| LMArena French | 1423 | 1282 |
| LMArena German | 1429 | 1251 |
| LMArena Japanese | 1411 | 1209 |
| LMArena Korean | 1393 | 1184 |
| LMArena Russian | 1420 | 1251 |
| LMArena Spanish | 1420 | 1261 |

## Instruction Following

- Gemini 3.1 Flash Lite: 72.9 (#126)
- GPT-4: 65.5 (#217)

| Benchmark | Gemini 3.1 Flash Lite | GPT-4 |
|---|---|---|
| LMArena Instruction Following | 1376 | 1241 |

## Long Context

- Gemini 3.1 Flash Lite: 42.5 (#117)
- GPT-4: 37.7 (#207)

| Benchmark | Gemini 3.1 Flash Lite | GPT-4 |
|---|---|---|
| LMArena Longer Query | 1394 | 1244 |

## Writing & Preference

- Gemini 3.1 Flash Lite: 61.0 (#90)
- GPT-4: 35.0 (#263)

| Benchmark | Gemini 3.1 Flash Lite | GPT-4 |
|---|---|---|
| LMArena Text | 1416 | 1263 |
| LMArena Creative Writing | 1401 | 1244 |
| LMArena Multi-Turn | 1417 | 1257 |
| EQ-Bench Creative Writing | — | 752 |

## FAQ

### Is Gemini 3.1 Flash Lite better than GPT-4?

Gemini 3.1 Flash Lite is the stronger model overall, scoring 40.4 to 28.9 on the Noometry Index.

### Which is cheaper, Gemini 3.1 Flash Lite or GPT-4?

Gemini 3.1 Flash Lite is cheaper. It lists at $0.25 per million input tokens and $1.50 per million output tokens; GPT-4 lists at $30 and $60.

### Is Gemini 3.1 Flash Lite or GPT-4 better for coding?

Gemini 3.1 Flash Lite scores higher on coding benchmarks: 37.5 versus 31.2 in the Noometry coding category.

### Which has the bigger context window?

Gemini 3.1 Flash Lite does, with 1.05M tokens against 8K.

### How many benchmarks do Gemini 3.1 Flash Lite and GPT-4 share?

25 benchmarks have published results for both models. Gemini 3.1 Flash Lite has 38 scored results on Noometry and GPT-4 has 38.