# Gemini 3.8 Flash vs GPT-4

> Gemini 3.8 Flash is the stronger model overall, scoring 61.8 to 29.1 on the Noometry Index.

- Canonical page: https://noometry.com/compare/gemini-3-8-flash-vs-gpt-4
- Last updated: 2026-10-10
- Shared benchmarks: 26

## Summary

- They share 26 benchmarks with published results for both. Gemini 3.8 Flash scores higher in 8 categories and GPT-4 in 0 categories; 8 gaps are clear of the uncertainty.
- The widest gap is in reasoning, where Gemini 3.8 Flash leads 76.9 to 17.8.
- The biggest single-benchmark swing is OTIS Mock AIME 2024-2025: 98.9% for Gemini 3.8 Flash and 1.1% for GPT-4.
- Gemini 3.8 Flash is cheaper at $0.75 / $3.75 per million input/output tokens, against $30 / $60 for GPT-4.
- Gemini 3.8 Flash accepts more context: 1.05M tokens versus 8K.

## Snapshot

| | Gemini 3.8 Flash | GPT-4 |
|---|---|---|
| Provider | Google | OpenAI |
| Noometry Index | 61.8 | 29.1 |
| Rank | 11 | 316 |
| Context | 1.05M | 8K |
| Input $/M | $0.75 | $30 |
| Output $/M | $3.75 | $60 |
| Weights | Proprietary | Proprietary |

## Coding

- Gemini 3.8 Flash: 59.2 (#15)
- GPT-4: 31.6 (#283)

| Benchmark | Gemini 3.8 Flash | GPT-4 |
|---|---|---|
| WeirdML | 84.8% | 12.4% |
| LMArena Coding | 1510 | 1254 |
| DeepSWE | 73.8% | — |
| FrontierCode | 41.2% | — |
| CursorBench | 39.6% | — |
| LMArena WebDev | 1584 | — |
| FrontierSWE | 19.6% | — |
| SciCode | 56.6% | — |
| BigCodeBench Instruct | — | 46% |
| BigCodeBench Complete | — | 57.2% |
| ALE-Bench | 1,270 | — |
| HumanEval+ | — | 79.3% |

## Agentic & Tool Use

- Gemini 3.8 Flash: 41.8 (#21)
- GPT-4: —

| Benchmark | Gemini 3.8 Flash | GPT-4 |
|---|---|---|
| APEX-Agents | 64.3% | — |
| Remote Labor Index | 5.8% | — |
| GDP.pdf | 23.4% | — |
| METR Time Horizons | — | 36.1% |
| Vending-Bench 2 | 5,094 | — |

## Reasoning

- Gemini 3.8 Flash: 76.9 (#5)
- GPT-4: 17.8 (#289)

| Benchmark | Gemini 3.8 Flash | GPT-4 |
|---|---|---|
| Chess Puzzles | 61% | 4% |
| LMArena Hard Prompts | 1508 | 1241 |
| Mystery Game Puzzles | 47% | 12% |
| DTBench | 95.7% | 62.7% |
| LMCA | 52.9% | 17.1% |
| Epoch Capabilities Index | 156.71 | 125.89 |
| ARC-AGI-2 | 89.2% | — |
| NYT Connections (extended) | 97.4% | — |
| ARC-AGI-1 | 98.5% | — |
| CritPt | 18.3% | — |
| Surface Evolver Bench | 76.9% | — |
| BIG-Bench Hard | — | 75.1% |
| ForecastBench | — | 57.8 |
| HellaSwag | — | 95.3% |
| WinoGrande | — | 87.5% |

## Math

- Gemini 3.8 Flash: 65.3 (#28)
- GPT-4: 10.8 (#309)

| Benchmark | Gemini 3.8 Flash | GPT-4 |
|---|---|---|
| OTIS Mock AIME 2024-2025 | 98.9% | 1.1% |
| LMArena Math | 1528 | 1269 |
| FrontierMath (Tiers 1-3) | 68.4% | — |
| FrontierMath Tier 4 | 22% | — |
| ProofBench | 48% | — |
| MATH Level 5 | — | 23% |
| GSM8K | — | 92% |

## Knowledge

- Gemini 3.8 Flash: 74.8 (#2)
- GPT-4: 18.4 (#282)

| Benchmark | Gemini 3.8 Flash | GPT-4 |
|---|---|---|
| GPQA Diamond | 95.4% | 35.7% |
| LMArena Expert | 1524 | 1211 |
| Humanity's Last Exam | 44.5% | — |
| SimpleQA Verified | 69.7% | — |
| MMLU | — | 86.4% |
| TriviaQA | — | 84.8% |

## Multimodal

- Gemini 3.8 Flash: 40.7 (#45)
- GPT-4: —

| Benchmark | Gemini 3.8 Flash | GPT-4 |
|---|---|---|
| LMArena Vision | 1314 | — |
| Blueprint-Bench 2 | 38.6% | — |
| Furniture Assembly | 31.7% | — |

## Multilingual

- Gemini 3.8 Flash: 58.0 (#5)
- GPT-4: 40.6 (#215)

| Benchmark | Gemini 3.8 Flash | GPT-4 |
|---|---|---|
| LMArena Non-English | 1491 | 1246 |
| LMArena Chinese | 1554 | 1242 |
| LMArena French | 1498 | 1283 |
| LMArena German | 1493 | 1251 |
| LMArena Japanese | 1502 | 1209 |
| LMArena Korean | 1459 | 1184 |
| LMArena Russian | 1515 | 1251 |
| LMArena Spanish | 1485 | 1261 |

## Instruction Following

- Gemini 3.8 Flash: 78.0 (#13)
- GPT-4: 65.3 (#222)

| Benchmark | Gemini 3.8 Flash | GPT-4 |
|---|---|---|
| LMArena Instruction Following | 1490 | 1241 |

## Long Context

- Gemini 3.8 Flash: 46.3 (#24)
- GPT-4: 37.7 (#212)

| Benchmark | Gemini 3.8 Flash | GPT-4 |
|---|---|---|
| LMArena Longer Query | 1508 | 1244 |

## Writing & Preference

- Gemini 3.8 Flash: 72.2 (#15)
- GPT-4: 34.9 (#268)

| Benchmark | Gemini 3.8 Flash | GPT-4 |
|---|---|---|
| LMArena Text | 1499 | 1263 |
| LMArena Creative Writing | 1492 | 1244 |
| EQ-Bench Creative Writing | 1748 | 752 |
| LMArena Multi-Turn | 1501 | 1257 |

## FAQ

### Is Gemini 3.8 Flash better than GPT-4?

Gemini 3.8 Flash is the stronger model overall, scoring 61.8 to 29.1 on the Noometry Index.

### Which is cheaper, Gemini 3.8 Flash or GPT-4?

Gemini 3.8 Flash is cheaper. It lists at $0.75 per million input tokens and $3.75 per million output tokens; GPT-4 lists at $30 and $60.

### Is Gemini 3.8 Flash or GPT-4 better for coding?

Gemini 3.8 Flash scores higher on coding benchmarks: 59.2 versus 31.6 in the Noometry coding category.

### Which has the bigger context window?

Gemini 3.8 Flash does, with 1.05M tokens against 8K.

### How many benchmarks do Gemini 3.8 Flash and GPT-4 share?

26 benchmarks have published results for both models. Gemini 3.8 Flash has 50 scored results on Noometry and GPT-4 has 38.