# Gemini 2.5 Pro vs GPT-5.2

> GPT-5.2 is the stronger model overall, scoring 54.1 to 44.9 on the Noometry Index.

- Canonical page: https://noometry.com/compare/gemini-2-5-pro-vs-gpt-5-2
- Last updated: 2026-10-10
- Shared benchmarks: 54

## Summary

- They share 54 benchmarks with published results for both. Gemini 2.5 Pro scores higher in 3 categories and GPT-5.2 in 7 categories; 9 gaps are clear of the uncertainty.
- The widest gap is in math, where GPT-5.2 leads 59.8 to 31.9.
- The biggest single-benchmark swing is ARC-AGI-2: 4.9% for Gemini 2.5 Pro and 52.9% for GPT-5.2.
- Gemini 2.5 Pro is cheaper at $1.25 / $10 per million input/output tokens, against $1.75 / $14 for GPT-5.2.
- Gemini 2.5 Pro accepts more context: 1.05M tokens versus 400K.

## Snapshot

| | Gemini 2.5 Pro | GPT-5.2 |
|---|---|---|
| Provider | Google | OpenAI |
| Noometry Index | 44.9 | 54.1 |
| Rank | 67 | 31 |
| Context | 1.05M | 400K |
| Input $/M | $1.25 | $1.75 |
| Output $/M | $10 | $14 |
| Weights | Proprietary | Proprietary |

## Coding

- Gemini 2.5 Pro: 42.8 (#86)
- GPT-5.2: 52.2 (#32)

| Benchmark | Gemini 2.5 Pro | GPT-5.2 |
|---|---|---|
| SWE-bench Verified | 57.6% | 73.8% |
| SWE-bench Verified (bash only) | 53.6% | 72.8% |
| LMArena WebDev | 1227 | 1415 |
| GSO | 3.9% | 27.4% |
| WeirdML | 54% | 72.2% |
| LMArena Coding | 1452 | 1447 |
| ALE-Bench | 785.52 | 1,294 |
| AlgoTune | 1.51 | 2.05 |
| Aider Polyglot | 83.1% | — |
| SWE-bench Multilingual | — | 66.7% |
| SciCode | 42.8% | — |
| LiveBench Coding | 85.9% | — |
| CadEval | 64% | — |

## Agentic & Tool Use

- Gemini 2.5 Pro: 29.1 (#81)
- GPT-5.2: 40.0 (#24)

| Benchmark | Gemini 2.5 Pro | GPT-5.2 |
|---|---|---|
| Terminal-Bench | 32.6% | 64.9% |
| GDPval | 23.3% | 49.7% |
| Remote Labor Index | 0.8% | 2.5% |
| τ²-bench Banking | 13.7% | 32.2% |
| DeepResearch Bench | 42.8% | 41.1% |
| LMArena Search | 1142 | 1207 |
| METR Time Horizons | 55.4% | 75.3% |
| Vending-Bench 2 | 573.64 | 3,591 |
| Berkeley Function Calling Leaderboard | — | 55.9% |
| TheAgentCompany | 30.3% | — |
| τ²-bench Airline | — | 83% |
| τ²-bench Retail | — | 81.6% |
| τ²-bench Telecom | — | 89.7% |
| BALROG | 43.3% | — |

## Reasoning

- Gemini 2.5 Pro: 28.8 (#85)
- GPT-5.2: 50.1 (#33)

| Benchmark | Gemini 2.5 Pro | GPT-5.2 |
|---|---|---|
| ARC-AGI-2 | 4.9% | 52.9% |
| SimpleBench | 62.4% | 45.8% |
| Kagi LLM Benchmark | 70.3% | 73.3% |
| ARC-AGI-1 | 41% | 86.2% |
| Chess Puzzles | 20% | 49% |
| EnigmaEval | 5.6% | 10.4% |
| LMArena Hard Prompts | 1455 | 1445 |
| DTBench | 82.4% | 90.9% |
| LMCA | 34.8% | 43.9% |
| Epoch Capabilities Index | 145.32 | 153.45 |
| ForecastBench | 61.3 | 60.1 |
| NYT Connections (extended) | — | 83.6% |
| CritPt | 2% | — |
| EBR-Bench | — | 23% |
| LiveBench Reasoning | 89.8% | — |
| Mystery Game Puzzles | — | 23% |
| LiveBench Data Analysis | 79.9% | — |
| LiveBench | 82.3% | — |

## Math

- Gemini 2.5 Pro: 31.9 (#191)
- GPT-5.2: 59.8 (#32)

| Benchmark | Gemini 2.5 Pro | GPT-5.2 |
|---|---|---|
| FrontierMath (Tiers 1-3) | 24.6% | 67.4% |
| FrontierMath Tier 4 | 0% | 31.7% |
| OTIS Mock AIME 2024-2025 | 84.7% | 96.1% |
| LMArena Math | 1450 | 1439 |
| FrontierMath (Feb 2025 set) | 14.1% | 40.7% |
| FrontierMath Tier 4 (v1) | 4.2% | 18.8% |
| MathArena Final-Answer Competitions | — | 72% |
| ProofBench | — | 15% |
| Omni-MATH | 41.6% | — |
| LiveBench Math | 90.2% | — |
| MATH Level 5 | 95.9% | — |

## Knowledge

- Gemini 2.5 Pro: 55.8 (#41)
- GPT-5.2: 59.2 (#29)

| Benchmark | Gemini 2.5 Pro | GPT-5.2 |
|---|---|---|
| GPQA Diamond | 85.3% | 91.4% |
| Humanity's Last Exam | 21.6% | 27.8% |
| Vectara Hallucination Rate | 7% | 8.4% |
| LMArena Expert | 1452 | 1445 |
| SimpleQA Verified | — | 37.1% |
| MMLU-Pro | 86.3% | — |
| Confabulations | 10.6% | — |
| GPQA (HELM) | 74.9% | — |

## Multimodal

- Gemini 2.5 Pro: 44.5 (#16)
- GPT-5.2: 50.9 (#7)

| Benchmark | Gemini 2.5 Pro | GPT-5.2 |
|---|---|---|
| LMArena Vision | 1263 | 1268 |
| VPCT | 48% | 84% |
| LMArena Document | 1421 | 1405 |
| GeoBench | 86% | — |
| Furniture Assembly | — | 38.3% |
| SpatialViz-Bench | 44.7% | — |

## Multilingual

- Gemini 2.5 Pro: 55.5 (#30)
- GPT-5.2: 53.6 (#63)

| Benchmark | Gemini 2.5 Pro | GPT-5.2 |
|---|---|---|
| LMArena Non-English | 1452 | 1425 |
| LMArena Chinese | 1507 | 1460 |
| LMArena French | 1473 | 1456 |
| LMArena German | 1487 | 1448 |
| LMArena Japanese | 1462 | 1421 |
| LMArena Korean | 1435 | 1390 |
| LMArena Russian | 1460 | 1440 |
| LMArena Spanish | 1473 | 1433 |

## Instruction Following

- Gemini 2.5 Pro: 75.2 (#71)
- GPT-5.2: 74.9 (#83)

| Benchmark | Gemini 2.5 Pro | GPT-5.2 |
|---|---|---|
| LMArena Instruction Following | 1436 | 1417 |
| LiveBench Instruction Following | 80.6% | — |
| IFEval | 84% | — |

## Long Context

- Gemini 2.5 Pro: 59.7 (#5)
- GPT-5.2: 44.0 (#74)

| Benchmark | Gemini 2.5 Pro | GPT-5.2 |
|---|---|---|
| LMArena Longer Query | 1449 | 1428 |
| Fiction.LiveBench | 91.7% | — |
| CL-bench | — | 18.2% |

## Writing & Preference

- Gemini 2.5 Pro: 63.9 (#60)
- GPT-5.2: 67.0 (#30)

| Benchmark | Gemini 2.5 Pro | GPT-5.2 |
|---|---|---|
| LMArena Text | 1458 | 1439 |
| LMArena Creative Writing | 1453 | 1402 |
| EQ-Bench Creative Writing | 1421 | 1703 |
| LMArena Multi-Turn | 1453 | 1458 |
| Short-Story Creative Writing | 83.8% | — |
| WildBench | 85.7% | — |
| LiveBench Language | 67.8% | — |

## FAQ

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

GPT-5.2 is the stronger model overall, scoring 54.1 to 44.9 on the Noometry Index.

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

Gemini 2.5 Pro is cheaper. It lists at $1.25 per million input tokens and $10 per million output tokens; GPT-5.2 lists at $1.75 and $14.

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

GPT-5.2 scores higher on coding benchmarks: 52.2 versus 42.8 in the Noometry coding category.

### Which has the bigger context window?

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

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

54 benchmarks have published results for both models. Gemini 2.5 Pro has 78 scored results on Noometry and GPT-5.2 has 67.