# Gemini 3 Pro vs GPT-5.6 Sol

> GPT-5.6 Sol is the stronger model overall, scoring 64.4 to 54.7 on the Noometry Index.

- Canonical page: https://noometry.com/compare/gemini-3-pro-vs-gpt-5-6-sol
- Last updated: 2026-10-10
- Shared benchmarks: 40

## Summary

- They share 40 benchmarks with published results for both. Gemini 3 Pro scores higher in 3 categories and GPT-5.6 Sol in 7 categories; 9 gaps are clear of the uncertainty.
- The widest gap is in math, where GPT-5.6 Sol leads 85.5 to 48.8.
- The biggest single-benchmark swing is ProofBench: 20% for Gemini 3 Pro and 83% for GPT-5.6 Sol.

## Snapshot

| | Gemini 3 Pro | GPT-5.6 Sol |
|---|---|---|
| Provider | Google | OpenAI |
| Noometry Index | 54.7 | 64.4 |
| Rank | 23 | 6 |
| Context | — | 1.05M |
| Input $/M | — | $4 |
| Output $/M | — | $20 |
| Weights | Proprietary | Proprietary |

## Coding

- Gemini 3 Pro: 52.1 (#33)
- GPT-5.6 Sol: 61.1 (#7)

| Benchmark | Gemini 3 Pro | GPT-5.6 Sol |
|---|---|---|
| LMArena WebDev | 1439 | 1619 |
| WeirdML | 69.9% | 89.4% |
| LMArena Coding | 1482 | 1493 |
| ALE-Bench | 1,177 | 2,177 |
| SWE-bench Verified | 72.9% | — |
| DeepSWE | — | 72.7% |
| FrontierCode | — | 47.5% |
| SWE-bench Verified (bash only) | 74.2% | — |
| CursorBench | — | 41.7% |
| SWE-bench Multilingual | 68.7% | — |
| FrontierSWE | — | 32.2% |
| SciCode | — | 57.1% |
| GSO | 18.6% | — |
| MirrorCode | — | 20% |
| AlgoTune | 1.83 | — |

## Agentic & Tool Use

- Gemini 3 Pro: 40.5 (#21)
- GPT-5.6 Sol: 50.2 (#7)

| Benchmark | Gemini 3 Pro | GPT-5.6 Sol |
|---|---|---|
| τ²-bench Banking | 18% | 46.9% |
| BALROG | 58.1% | 60% |
| LMArena Search | 1207 | 1257 |
| Vending-Bench 2 | 5,478 | 9,619 |
| Terminal-Bench | 69.4% | — |
| APEX-Agents | — | 51.4% |
| Berkeley Function Calling Leaderboard | 72.5% | — |
| OSWorld 2.0 | — | 27.3% |
| GDPval | 40.3% | — |
| Remote Labor Index | 1.3% | — |
| τ²-bench Airline | 80.5% | — |
| τ²-bench Retail | 75.9% | — |
| τ²-bench Telecom | 91% | — |
| DeepResearch Bench | 46.3% | — |
| PostTrainBench | — | 36.2% |
| GBAEval | — | 52.6% |
| GDP.pdf | — | 30.7% |
| METR Time Horizons | 71% | — |

## Reasoning

- Gemini 3 Pro: 52.5 (#28)
- GPT-5.6 Sol: 74.7 (#6)

| Benchmark | Gemini 3 Pro | GPT-5.6 Sol |
|---|---|---|
| ARC-AGI-2 | 31.1% | 92.5% |
| SimpleBench | 76.4% | 71.7% |
| Kagi LLM Benchmark | 80.1% | 67% |
| NYT Connections (extended) | 94.4% | 93.8% |
| ARC-AGI-1 | 75% | 97.5% |
| CritPt | 6.9% | 32.3% |
| Chess Puzzles | 31% | 64% |
| EnigmaEval | 18.2% | 37.1% |
| LMArena Hard Prompts | 1481 | 1481 |
| Epoch Capabilities Index | 152.96 | 161.99 |
| EBR-Bench | — | 44.8% |
| Mystery Game Puzzles | — | 58% |
| DTBench | — | 96% |
| LMCA | — | 59.2% |
| Surface Evolver Bench | — | 93.1% |
| Bench to the Future 3 | — | 0.14 |
| ForecastBench | 61.2 | — |

## Math

- Gemini 3 Pro: 48.8 (#50)
- GPT-5.6 Sol: 85.5 (#7)

| Benchmark | Gemini 3 Pro | GPT-5.6 Sol |
|---|---|---|
| OTIS Mock AIME 2024-2025 | 91.4% | 100% |
| ProofBench | 20% | 83% |
| LMArena Math | 1475 | 1472 |
| FrontierMath (Tiers 1-3) | — | 89.1% |
| FrontierMath Tier 4 | — | 82.9% |
| MathArena Final-Answer Competitions | 67% | — |
| Omni-MATH | 55.5% | — |
| FrontierMath (Feb 2025 set) | 37.6% | — |
| FrontierMath Erdős | — | 0% |
| FrontierMath Tier 4 (v1) | 18.8% | — |

## Knowledge

- Gemini 3 Pro: 64.3 (#15)
- GPT-5.6 Sol: 64.3 (#16)

| Benchmark | Gemini 3 Pro | GPT-5.6 Sol |
|---|---|---|
| GPQA Diamond | 92.6% | 93.5% |
| Vectara Hallucination Rate | 13.6% | 12.4% |
| LMArena Expert | 1475 | 1517 |
| Humanity's Last Exam | 37.5% | — |
| SimpleQA Verified | — | 69.7% |
| MMLU-Pro | 90.3% | — |
| GPQA (HELM) | 80.3% | — |

## Multimodal

- Gemini 3 Pro: 56.9 (#1)
- GPT-5.6 Sol: 48.9 (#8)

| Benchmark | Gemini 3 Pro | GPT-5.6 Sol |
|---|---|---|
| LMArena Vision | 1305 | 1281 |
| LMArena Document | 1434 | 1483 |
| GeoBench | 84% | — |
| VPCT | 91% | — |
| Blueprint-Bench 2 | — | 33.6% |
| Furniture Assembly | — | 56.7% |

## Multilingual

- Gemini 3 Pro: 57.1 (#14)
- GPT-5.6 Sol: 55.4 (#32)

| Benchmark | Gemini 3 Pro | GPT-5.6 Sol |
|---|---|---|
| LMArena Non-English | 1474 | 1450 |
| LMArena Chinese | 1522 | 1523 |
| LMArena French | 1492 | 1475 |
| LMArena German | 1517 | 1476 |
| LMArena Japanese | 1513 | 1475 |
| LMArena Korean | 1449 | 1434 |
| LMArena Russian | 1493 | 1466 |
| LMArena Spanish | 1468 | 1449 |

## Instruction Following

- Gemini 3 Pro: 76.4 (#42)
- GPT-5.6 Sol: 77.8 (#16)

| Benchmark | Gemini 3 Pro | GPT-5.6 Sol |
|---|---|---|
| LMArena Instruction Following | 1458 | 1480 |
| IFEval | 87.7% | — |

## Long Context

- Gemini 3 Pro: 44.0 (#75)
- GPT-5.6 Sol: 45.3 (#42)

| Benchmark | Gemini 3 Pro | GPT-5.6 Sol |
|---|---|---|
| LMArena Longer Query | 1472 | 1479 |
| CL-bench | 15.8% | — |

## Writing & Preference

- Gemini 3 Pro: 66.6 (#34)
- GPT-5.6 Sol: 73.5 (#12)

| Benchmark | Gemini 3 Pro | GPT-5.6 Sol |
|---|---|---|
| LMArena Text | 1479 | 1456 |
| LMArena Creative Writing | 1482 | 1450 |
| EQ-Bench Creative Writing | 1525 | 1972 |
| LMArena Multi-Turn | 1484 | 1456 |
| WildBench | 85.9% | — |
| EQ-Bench 4 | — | 1250 |

## FAQ

### Is Gemini 3 Pro better than GPT-5.6 Sol?

GPT-5.6 Sol is the stronger model overall, scoring 64.4 to 54.7 on the Noometry Index.

### Is Gemini 3 Pro or GPT-5.6 Sol better for coding?

GPT-5.6 Sol scores higher on coding benchmarks: 61.1 versus 52.1 in the Noometry coding category.

### How many benchmarks do Gemini 3 Pro and GPT-5.6 Sol share?

40 benchmarks have published results for both models. Gemini 3 Pro has 67 scored results on Noometry and GPT-5.6 Sol has 64.