# Gemini 2.5 Flash vs GPT-5.2

> GPT-5.2 is the stronger model overall, scoring 54.1 to 39.3 on the Noometry Index. Gemini 2.5 Flash costs 5.7× less per token, which makes it the better buy when GPT-5.2's lead doesn't matter for your workload.

- Canonical page: https://noometry.com/compare/gemini-2-5-flash-vs-gpt-5-2
- Last updated: 2026-10-10
- Shared benchmarks: 40

## Summary

- They share 40 benchmarks with published results for both. Gemini 2.5 Flash scores higher in 2 categories and GPT-5.2 in 8 categories; 10 gaps are clear of the uncertainty.
- The widest gap is in reasoning, where GPT-5.2 leads 50.2 to 18.1.
- The biggest single-benchmark swing is ARC-AGI-1: 33.3% for Gemini 2.5 Flash and 86.2% for GPT-5.2.
- Gemini 2.5 Flash is cheaper at $0.30 / $2.50 per million input/output tokens, against $1.75 / $14 for GPT-5.2.
- Gemini 2.5 Flash accepts more context: 1.05M tokens versus 400K.

## Snapshot

| | Gemini 2.5 Flash | GPT-5.2 |
|---|---|---|
| Provider | Google | OpenAI |
| Noometry Index | 39.3 | 54.1 |
| Rank | 170 | 34 |
| Context | 1.05M | 400K |
| Input $/M | $0.30 | $1.75 |
| Output $/M | $2.50 | $14 |
| Weights | Proprietary | Proprietary |

## Coding

- Gemini 2.5 Flash: 35.8 (#220)
- GPT-5.2: 51.6 (#37)

| Benchmark | Gemini 2.5 Flash | GPT-5.2 |
|---|---|---|
| SWE-bench Verified (bash only) | 28.7% | 72.8% |
| WeirdML | 41.9% | 72.2% |
| LMArena Coding | 1424 | 1447 |
| ALE-Bench | 661.88 | 1,294 |
| SWE-bench Verified | — | 73.8% |
| Aider Polyglot | 55.1% | — |
| LMArena WebDev | — | 1416 |
| SWE-bench Multilingual | — | 66.7% |
| GSO | — | 27.4% |
| AlgoTune | — | 2.05 |

## Agentic & Tool Use

- Gemini 2.5 Flash: 30.8 (#74)
- GPT-5.2: 40.2 (#24)

| Benchmark | Gemini 2.5 Flash | GPT-5.2 |
|---|---|---|
| Terminal-Bench | 17.1% | 64.9% |
| Berkeley Function Calling Leaderboard | 56.2% | 55.9% |
| Vending-Bench 2 | 548.84 | 3,591 |
| GDPval | — | 49.7% |
| Remote Labor Index | — | 2.5% |
| TheAgentCompany | 41.1% | — |
| τ²-bench Airline | — | 83% |
| τ²-bench Banking | — | 32.2% |
| τ²-bench Retail | — | 81.6% |
| τ²-bench Telecom | — | 89.7% |
| DeepResearch Bench | — | 41.1% |
| BALROG | 33.5% | — |
| LMArena Search | — | 1207 |
| METR Time Horizons | — | 75.3% |

## Reasoning

- Gemini 2.5 Flash: 18.1 (#286)
- GPT-5.2: 50.2 (#35)

| Benchmark | Gemini 2.5 Flash | GPT-5.2 |
|---|---|---|
| ARC-AGI-2 | 2.5% | 52.9% |
| SimpleBench | 41.2% | 45.8% |
| Kagi LLM Benchmark | 56.8% | 73.3% |
| ARC-AGI-1 | 33.3% | 86.2% |
| EnigmaEval | 2.7% | 10.4% |
| LMArena Hard Prompts | 1422 | 1445 |
| DTBench | 76.5% | 90.9% |
| LMCA | 27.5% | 43.9% |
| Epoch Capabilities Index | 143.03 | 153.45 |
| ForecastBench | 60.6 | 60.1 |
| NYT Connections (extended) | — | 83.6% |
| CritPt | 1.1% | — |
| Chess Puzzles | — | 49% |
| EBR-Bench | — | 23% |
| Mystery Game Puzzles | — | 23% |

## Math

- Gemini 2.5 Flash: 39.9 (#98)
- GPT-5.2: 60.0 (#38)

| Benchmark | Gemini 2.5 Flash | GPT-5.2 |
|---|---|---|
| OTIS Mock AIME 2024-2025 | 73.1% | 96.1% |
| LMArena Math | 1415 | 1440 |
| FrontierMath (Feb 2025 set) | 4.8% | 40.7% |
| FrontierMath Tier 4 (v1) | 4.2% | 18.8% |
| FrontierMath (Tiers 1-3) | — | 67.4% |
| FrontierMath Tier 4 | — | 31.7% |
| MathArena Final-Answer Competitions | — | 72% |
| ProofBench | — | 15% |
| Omni-MATH | 38.5% | — |

## Knowledge

- Gemini 2.5 Flash: 36.4 (#168)
- GPT-5.2: 59.3 (#32)

| Benchmark | Gemini 2.5 Flash | GPT-5.2 |
|---|---|---|
| Humanity's Last Exam | 12.1% | 27.8% |
| Vectara Hallucination Rate | 7.8% | 8.4% |
| LMArena Expert | 1426 | 1445 |
| GPQA Diamond | — | 91.4% |
| SimpleQA Verified | — | 37.1% |
| MMLU-Pro | 63.9% | — |
| Confabulations | 16.8% | — |
| GPQA (HELM) | 39% | — |

## Multimodal

- Gemini 2.5 Flash: 41.8 (#32)
- GPT-5.2: 51.3 (#7)

| Benchmark | Gemini 2.5 Flash | GPT-5.2 |
|---|---|---|
| LMArena Vision | 1253 | 1268 |
| VPCT | 46.2% | 84% |
| GeoBench | 76% | — |
| Furniture Assembly | — | 38.3% |
| LMArena Document | — | 1405 |
| SpatialViz-Bench | 36.9% | — |

## Multilingual

- Gemini 2.5 Flash: 52.3 (#88)
- GPT-5.2: 53.4 (#67)

| Benchmark | Gemini 2.5 Flash | GPT-5.2 |
|---|---|---|
| LMArena Non-English | 1409 | 1425 |
| LMArena Chinese | 1450 | 1460 |
| LMArena French | 1433 | 1455 |
| LMArena German | 1418 | 1448 |
| LMArena Japanese | 1405 | 1420 |
| LMArena Korean | 1385 | 1392 |
| LMArena Russian | 1415 | 1440 |
| LMArena Spanish | 1421 | 1433 |

## Instruction Following

- Gemini 2.5 Flash: 75.7 (#54)
- GPT-5.2: 74.7 (#89)

| Benchmark | Gemini 2.5 Flash | GPT-5.2 |
|---|---|---|
| LMArena Instruction Following | 1405 | 1417 |
| IFEval | 89.8% | — |

## Long Context

- Gemini 2.5 Flash: 47.5 (#17)
- GPT-5.2: 44.0 (#78)

| Benchmark | Gemini 2.5 Flash | GPT-5.2 |
|---|---|---|
| LMArena Longer Query | 1419 | 1428 |
| Fiction.LiveBench | 77.8% | — |
| CL-bench | — | 18.2% |

## Writing & Preference

- Gemini 2.5 Flash: 53.8 (#157)
- GPT-5.2: 66.8 (#32)

| Benchmark | Gemini 2.5 Flash | GPT-5.2 |
|---|---|---|
| LMArena Text | 1417 | 1439 |
| LMArena Creative Writing | 1400 | 1401 |
| EQ-Bench Creative Writing | 1137 | 1703 |
| LMArena Multi-Turn | 1408 | 1458 |
| Short-Story Creative Writing | 76.5% | — |
| WildBench | 81.7% | — |

## FAQ

### Is Gemini 2.5 Flash better than GPT-5.2?

GPT-5.2 is the stronger model overall, scoring 54.1 to 39.3 on the Noometry Index. Gemini 2.5 Flash costs 5.7× less per token, which makes it the better buy when GPT-5.2's lead doesn't matter for your workload.

### Which is cheaper, Gemini 2.5 Flash or GPT-5.2?

Gemini 2.5 Flash is cheaper. It lists at $0.30 per million input tokens and $2.50 per million output tokens; GPT-5.2 lists at $1.75 and $14.

### Is Gemini 2.5 Flash or GPT-5.2 better for coding?

GPT-5.2 scores higher on coding benchmarks: 51.6 versus 35.8 in the Noometry coding category.

### Which has the bigger context window?

Gemini 2.5 Flash does, with 1.05M tokens against 400K.

### How many benchmarks do Gemini 2.5 Flash and GPT-5.2 share?

40 benchmarks have published results for both models. Gemini 2.5 Flash has 54 scored results on Noometry and GPT-5.2 has 67.