# Gemini 3 Flash Preview vs GPT-4o

> Gemini 3 Flash Preview is the stronger model overall, scoring 52.3 to 28.6 on the Noometry Index.

- Canonical page: https://noometry.com/compare/gemini-3-flash-preview-vs-gpt-4o
- Last updated: 2026-10-11
- Shared benchmarks: 40

## Summary

- They share 40 benchmarks with published results for both. Gemini 3 Flash Preview scores higher in 10 categories and GPT-4o in 0 categories; 10 gaps are clear of the uncertainty.
- The widest gap is in math, where Gemini 3 Flash Preview leads 51.7 to 10.6.
- The biggest single-benchmark swing is OTIS Mock AIME 2024-2025: 95.6% for Gemini 3 Flash Preview and 6.4% for GPT-4o.
- Gemini 3 Flash Preview is cheaper at $0.50 / $3 per million input/output tokens, against $2.50 / $10 for GPT-4o.
- Gemini 3 Flash Preview accepts more context: 1.05M tokens versus 128K.

## Snapshot

| | Gemini 3 Flash Preview | GPT-4o |
|---|---|---|
| Provider | Google | OpenAI |
| Noometry Index | 52.3 | 28.6 |
| Rank | 40 | 324 |
| Context | 1.05M | 128K |
| Input $/M | $0.50 | $2.50 |
| Output $/M | $3 | $10 |
| Weights | Proprietary | Proprietary |

## Coding

- Gemini 3 Flash Preview: 50.9 (#42)
- GPT-4o: 24.8 (#328)

| Benchmark | Gemini 3 Flash Preview | GPT-4o |
|---|---|---|
| SWE-bench Verified | 75.4% | 31% |
| SWE-bench Verified (bash only) | 75.8% | 21.6% |
| GSO | 9.8% | 0% |
| WeirdML | 61.6% | 25.1% |
| LMArena Coding | 1460 | 1297 |
| Aider Polyglot | — | 45.3% |
| LMArena WebDev | 1439 | — |
| SWE-bench Multilingual | 72.7% | — |
| BigCodeBench Instruct | — | 51.1% |
| LiveBench Coding | — | 51.4% |
| BigCodeBench Complete | — | 61.1% |
| CadEval | — | 26% |
| ALE-Bench | 1,367 | — |
| HumanEval+ | — | 87.2% |
| MBPP+ | — | 72.2% |

## Agentic & Tool Use

- Gemini 3 Flash Preview: 38.7 (#29)
- GPT-4o: 21.0 (#141)

| Benchmark | Gemini 3 Flash Preview | GPT-4o |
|---|---|---|
| BALROG | 48.1% | 32.3% |
| LMArena Search | 1198 | 1006 |
| Terminal-Bench | 64.3% | — |
| GDPval | — | 9.9% |
| TheAgentCompany | — | 8.6% |
| τ²-bench Airline | 82.5% | — |
| τ²-bench Banking | 27.3% | — |
| τ²-bench Retail | 76.8% | — |
| τ²-bench Telecom | 91.2% | — |
| Cybench | — | 12.5% |
| DeepResearch Bench | 49.8% | — |
| GDP.pdf | 10% | — |
| METR Time Horizons | — | 40.8% |
| Vending-Bench 2 | 3,635 | — |

## Reasoning

- Gemini 3 Flash Preview: 49.2 (#37)
- GPT-4o: 9.4 (#343)

| Benchmark | Gemini 3 Flash Preview | GPT-4o |
|---|---|---|
| ARC-AGI-2 | 33.6% | 0% |
| SimpleBench | 61.1% | 17.8% |
| ARC-AGI-1 | 84.7% | 4.5% |
| Chess Puzzles | 40% | 13% |
| LMArena Hard Prompts | 1465 | 1281 |
| DTBench | 89.1% | 64.5% |
| LMCA | 43.1% | 16.6% |
| Epoch Capabilities Index | 151.8 | 128.97 |
| ForecastBench | 58.5 | 57.7 |
| NYT Connections (extended) | 83.1% | — |
| CritPt | — | 0% |
| EnigmaEval | — | 0.8% |
| LiveBench Reasoning | — | 55.8% |
| Mystery Game Puzzles | 26% | — |
| LiveBench Data Analysis | — | 60.9% |
| LiveBench | — | 55.3% |

## Math

- Gemini 3 Flash Preview: 51.7 (#55)
- GPT-4o: 10.6 (#312)

| Benchmark | Gemini 3 Flash Preview | GPT-4o |
|---|---|---|
| FrontierMath (Tiers 1-3) | 51.2% | 0.4% |
| OTIS Mock AIME 2024-2025 | 95.6% | 6.4% |
| LMArena Math | 1473 | 1285 |
| FrontierMath (Feb 2025 set) | 35.6% | 0.3% |
| FrontierMath Tier 4 | 17.1% | — |
| MathArena Final-Answer Competitions | 67.6% | — |
| ProofBench | 15% | — |
| Omni-MATH | — | 29.3% |
| LiveBench Math | — | 49.5% |
| MATH Level 5 | — | 53.3% |
| FrontierMath Tier 4 (v1) | 4.2% | — |

## Knowledge

- Gemini 3 Flash Preview: 58.8 (#33)
- GPT-4o: 28.8 (#242)

| Benchmark | Gemini 3 Flash Preview | GPT-4o |
|---|---|---|
| GPQA Diamond | 89.4% | 49.2% |
| SimpleQA Verified | 66.8% | 26% |
| Vectara Hallucination Rate | 13.5% | 9.6% |
| LMArena Expert | 1462 | 1250 |
| Humanity's Last Exam | — | 2.7% |
| MMLU-Pro | — | 71.3% |
| Confabulations | — | 15.3% |
| GPQA (HELM) | — | 52% |
| MMLU | — | 88.1% |

## Multimodal

- Gemini 3 Flash Preview: 45.5 (#16)
- GPT-4o: 34.5 (#91)

| Benchmark | Gemini 3 Flash Preview | GPT-4o |
|---|---|---|
| LMArena Vision | 1285 | 1137 |
| GeoBench | 88% | 71% |
| VPCT | 72.6% | 40% |
| Video-MME | — | 71.9% |
| Blueprint-Bench 2 | 0% | — |
| LMArena Document | 1413 | — |
| ScienceQA | — | 88.5% |

## Multilingual

- Gemini 3 Flash Preview: 55.7 (#27)
- GPT-4o: 43.2 (#186)

| Benchmark | Gemini 3 Flash Preview | GPT-4o |
|---|---|---|
| LMArena Non-English | 1458 | 1283 |
| LMArena Chinese | 1511 | 1277 |
| LMArena French | 1477 | 1304 |
| LMArena German | 1497 | 1282 |
| LMArena Japanese | 1489 | 1257 |
| LMArena Korean | 1443 | 1234 |
| LMArena Russian | 1480 | 1286 |
| LMArena Spanish | 1469 | 1292 |

## Instruction Following

- Gemini 3 Flash Preview: 75.7 (#56)
- GPT-4o: 66.6 (#207)

| Benchmark | Gemini 3 Flash Preview | GPT-4o |
|---|---|---|
| LMArena Instruction Following | 1437 | 1278 |
| LiveBench Instruction Following | — | 68.6% |
| IFEval | — | 81.7% |

## Long Context

- Gemini 3 Flash Preview: 44.4 (#67)
- GPT-4o: 39.4 (#179)

| Benchmark | Gemini 3 Flash Preview | GPT-4o |
|---|---|---|
| LMArena Longer Query | 1452 | 1289 |
| Fiction.LiveBench | — | 66.7% |

## Writing & Preference

- Gemini 3 Flash Preview: 65.5 (#45)
- GPT-4o: 52.6 (#166)

| Benchmark | Gemini 3 Flash Preview | GPT-4o |
|---|---|---|
| LMArena Text | 1466 | 1300 |
| LMArena Creative Writing | 1457 | 1292 |
| LMArena Multi-Turn | 1471 | 1302 |
| Short-Story Creative Writing | — | 81.8% |
| WildBench | — | 82.8% |
| LiveBench Language | — | 47.6% |

## FAQ

### Is Gemini 3 Flash Preview better than GPT-4o?

Gemini 3 Flash Preview is the stronger model overall, scoring 52.3 to 28.6 on the Noometry Index.

### Which is cheaper, Gemini 3 Flash Preview or GPT-4o?

Gemini 3 Flash Preview is cheaper. It lists at $0.50 per million input tokens and $3 per million output tokens; GPT-4o lists at $2.50 and $10.

### Is Gemini 3 Flash Preview or GPT-4o better for coding?

Gemini 3 Flash Preview scores higher on coding benchmarks: 50.9 versus 24.8 in the Noometry coding category.

### Which has the bigger context window?

Gemini 3 Flash Preview does, with 1.05M tokens against 128K.

### How many benchmarks do Gemini 3 Flash Preview and GPT-4o share?

40 benchmarks have published results for both models. Gemini 3 Flash Preview has 59 scored results on Noometry and GPT-4o has 72.