# Gemini 2.5 Pro vs Gemini 3.1 Pro Preview

> Gemini 3.1 Pro Preview is the stronger model overall, scoring 56.7 to 44.9 on the Noometry Index.

- Canonical page: https://noometry.com/compare/gemini-2-5-pro-vs-gemini-3-1-pro-preview
- Last updated: 2026-10-10
- Shared benchmarks: 52

## Summary

- They share 52 benchmarks with published results for both. Gemini 2.5 Pro scores higher in 3 categories and Gemini 3.1 Pro Preview in 7 categories; 9 gaps are clear of the uncertainty.
- The widest gap is in reasoning, where Gemini 3.1 Pro Preview leads 71.7 to 28.8.
- The biggest single-benchmark swing is ARC-AGI-2: 4.9% for Gemini 2.5 Pro and 77.1% for Gemini 3.1 Pro Preview.
- Gemini 2.5 Pro is cheaper at $1.25 / $10 per million input/output tokens, against $2 / $12 for Gemini 3.1 Pro Preview.

## Snapshot

| | Gemini 2.5 Pro | Gemini 3.1 Pro Preview |
|---|---|---|
| Provider | Google | Google |
| Noometry Index | 44.9 | 56.7 |
| Rank | 67 | 20 |
| Context | 1.05M | 1.05M |
| Input $/M | $1.25 | $2 |
| Output $/M | $10 | $12 |
| Weights | Proprietary | Proprietary |

## Coding

- Gemini 2.5 Pro: 42.8 (#86)
- Gemini 3.1 Pro Preview: 42.7 (#88)

| Benchmark | Gemini 2.5 Pro | Gemini 3.1 Pro Preview |
|---|---|---|
| SWE-bench Verified | 57.6% | 75.6% |
| LMArena WebDev | 1227 | 1446 |
| SciCode | 42.8% | 58.9% |
| GSO | 3.9% | 22.6% |
| WeirdML | 54% | 72.1% |
| LMArena Coding | 1452 | 1483 |
| ALE-Bench | 785.52 | 1,161 |
| AlgoTune | 1.51 | 2.02 |
| DeepSWE | — | 11.7% |
| SWE-bench Verified (bash only) | 53.6% | — |
| Aider Polyglot | 83.1% | — |
| LiveBench Coding | 85.9% | — |
| MirrorCode | — | 8.9% |
| CadEval | 64% | — |

## Agentic & Tool Use

- Gemini 2.5 Pro: 29.1 (#81)
- Gemini 3.1 Pro Preview: 37.7 (#33)

| Benchmark | Gemini 2.5 Pro | Gemini 3.1 Pro Preview |
|---|---|---|
| Terminal-Bench | 32.6% | 80.2% |
| τ²-bench Banking | 13.7% | 26% |
| DeepResearch Bench | 42.8% | 47.8% |
| BALROG | 43.3% | 57% |
| LMArena Search | 1142 | 1211 |
| METR Time Horizons | 55.4% | 77% |
| Vending-Bench 2 | 573.64 | 3,774 |
| APEX-Agents | — | 35.3% |
| GDPval | 23.3% | — |
| Remote Labor Index | 0.8% | — |
| TheAgentCompany | 30.3% | — |
| PostTrainBench | — | 22% |
| ExploitBench | — | 26.1% |
| GBAEval | — | 0.8% |
| GDP.pdf | — | 17% |

## Reasoning

- Gemini 2.5 Pro: 28.8 (#85)
- Gemini 3.1 Pro Preview: 71.7 (#9)

| Benchmark | Gemini 2.5 Pro | Gemini 3.1 Pro Preview |
|---|---|---|
| ARC-AGI-2 | 4.9% | 77.1% |
| SimpleBench | 62.4% | 79.6% |
| ARC-AGI-1 | 41% | 98% |
| CritPt | 2% | 17.7% |
| Chess Puzzles | 20% | 55% |
| EnigmaEval | 5.6% | 36.8% |
| LMArena Hard Prompts | 1455 | 1484 |
| DTBench | 82.4% | 97.1% |
| LMCA | 34.8% | 53.8% |
| Epoch Capabilities Index | 145.32 | 154.92 |
| ForecastBench | 61.3 | 59 |
| Kagi LLM Benchmark | 70.3% | — |
| NYT Connections (extended) | — | 97.4% |
| Thematic Generalization | — | 79.4% |
| EBR-Bench | — | 14.3% |
| LiveBench Reasoning | 89.8% | — |
| Mystery Game Puzzles | — | 34% |
| LiveBench Data Analysis | 79.9% | — |
| LiveBench | 82.3% | — |

## Math

- Gemini 2.5 Pro: 31.9 (#191)
- Gemini 3.1 Pro Preview: 61.8 (#28)

| Benchmark | Gemini 2.5 Pro | Gemini 3.1 Pro Preview |
|---|---|---|
| FrontierMath (Tiers 1-3) | 24.6% | 59.6% |
| FrontierMath Tier 4 | 0% | 26.8% |
| OTIS Mock AIME 2024-2025 | 84.7% | 95.6% |
| LMArena Math | 1450 | 1485 |
| FrontierMath (Feb 2025 set) | 14.1% | 36.9% |
| FrontierMath Tier 4 (v1) | 4.2% | 16.7% |
| MathArena Final-Answer Competitions | — | 86.5% |
| ProofBench | — | 26% |
| Omni-MATH | 41.6% | — |
| LiveBench Math | 90.2% | — |
| MATH Level 5 | 95.9% | — |

## Knowledge

- Gemini 2.5 Pro: 55.8 (#41)
- Gemini 3.1 Pro Preview: 71.8 (#3)

| Benchmark | Gemini 2.5 Pro | Gemini 3.1 Pro Preview |
|---|---|---|
| GPQA Diamond | 85.3% | 94.4% |
| Humanity's Last Exam | 21.6% | 46.4% |
| Vectara Hallucination Rate | 7% | 10.4% |
| LMArena Expert | 1452 | 1484 |
| SimpleQA Verified | — | 73.5% |
| MMLU-Pro | 86.3% | — |
| Confabulations | 10.6% | — |
| GPQA (HELM) | 74.9% | — |

## Multimodal

- Gemini 2.5 Pro: 44.5 (#16)
- Gemini 3.1 Pro Preview: 38.1 (#60)

| Benchmark | Gemini 2.5 Pro | Gemini 3.1 Pro Preview |
|---|---|---|
| LMArena Vision | 1263 | 1296 |
| LMArena Document | 1421 | 1444 |
| GeoBench | 86% | — |
| VPCT | 48% | — |
| Blueprint-Bench 2 | — | 26.5% |
| Furniture Assembly | — | 26.7% |
| SpatialViz-Bench | 44.7% | — |

## Multilingual

- Gemini 2.5 Pro: 55.5 (#30)
- Gemini 3.1 Pro Preview: 57.1 (#11)

| Benchmark | Gemini 2.5 Pro | Gemini 3.1 Pro Preview |
|---|---|---|
| LMArena Non-English | 1452 | 1475 |
| LMArena Chinese | 1507 | 1528 |
| LMArena French | 1473 | 1488 |
| LMArena German | 1487 | 1491 |
| LMArena Japanese | 1462 | 1494 |
| LMArena Korean | 1435 | 1458 |
| LMArena Russian | 1460 | 1496 |
| LMArena Spanish | 1473 | 1480 |

## Instruction Following

- Gemini 2.5 Pro: 75.2 (#71)
- Gemini 3.1 Pro Preview: 77.1 (#31)

| Benchmark | Gemini 2.5 Pro | Gemini 3.1 Pro Preview |
|---|---|---|
| LMArena Instruction Following | 1436 | 1465 |
| LiveBench Instruction Following | 80.6% | — |
| IFEval | 84% | — |

## Long Context

- Gemini 2.5 Pro: 59.7 (#5)
- Gemini 3.1 Pro Preview: 47.3 (#17)

| Benchmark | Gemini 2.5 Pro | Gemini 3.1 Pro Preview |
|---|---|---|
| LMArena Longer Query | 1449 | 1482 |
| Fiction.LiveBench | 91.7% | — |
| CL-bench | — | 20.8% |
| CL-bench Life | — | 16.9% |

## Writing & Preference

- Gemini 2.5 Pro: 63.9 (#60)
- Gemini 3.1 Pro Preview: 66.3 (#36)

| Benchmark | Gemini 2.5 Pro | Gemini 3.1 Pro Preview |
|---|---|---|
| LMArena Text | 1458 | 1480 |
| LMArena Creative Writing | 1453 | 1482 |
| EQ-Bench Creative Writing | 1421 | 1491 |
| LMArena Multi-Turn | 1453 | 1486 |
| Short-Story Creative Writing | 83.8% | — |
| WildBench | 85.7% | — |
| EQ-Bench 4 | — | 1142 |
| LiveBench Language | 67.8% | — |

## FAQ

### Is Gemini 2.5 Pro better than Gemini 3.1 Pro Preview?

Gemini 3.1 Pro Preview is the stronger model overall, scoring 56.7 to 44.9 on the Noometry Index.

### Which is cheaper, Gemini 2.5 Pro or Gemini 3.1 Pro Preview?

Gemini 2.5 Pro is cheaper. It lists at $1.25 per million input tokens and $10 per million output tokens; Gemini 3.1 Pro Preview lists at $2 and $12.

### Is Gemini 2.5 Pro or Gemini 3.1 Pro Preview better for coding?

They score almost the same on coding (42.8 vs 42.7); test both on your own repository before choosing.

### Which has the bigger context window?

Both accept 1.05M tokens.

### How many benchmarks do Gemini 2.5 Pro and Gemini 3.1 Pro Preview share?

52 benchmarks have published results for both models. Gemini 2.5 Pro has 78 scored results on Noometry and Gemini 3.1 Pro Preview has 71.